PRATHYUSHA

ENGINEERING COLLEGE
DEPARTMENT OF COMPUTER SCIENCE AND
ENGINEERING

REGULATION 2017
IV YEAR - VII SEMESTER

CS8082 MACHINE LEARNING TECHNIQUES

CS8082 - MACHINE LEARNING TECHNIQUES

OBJECTIVES:

• To understand the need for machine learning for various problem solving
• To study the various supervised, semi-supervised and unsupervised learning
algorithms in machine learning
• To understand the latest trends in machine learning
• To design appropriate machine learning algorithms for problem solving

UNIT I INTRODUCTION 9

Learning Problems – Perspectives and Issues – Concept Learning – Version Spaces and
Candidate Eliminations – Inductive bias – Decision Tree learning – Representation –
Algorithm – Heuristic Space Search.

UNIT II NEURAL NETWORKS AND GENETIC

ALGORITHMS 9

Neural Network Representation – Problems – Perceptrons – Multilayer Networks and Back

Propagation Algorithms – Advanced Topics – Genetic Algorithms – Hypothesis Space
Search – Genetic Programming – Models of Evaluation and Learning.

UNIT III BAYESIAN AND COMPUTATIONAL

LEARNING 9

Bayes Theorem – Concept Learning – Maximum Likelihood – Minimum Description Length

Principle – Bayes Optimal Classifier – Gibbs Algorithm – Naïve Bayes Classifier – Bayesian
Belief Network – EM Algorithm – Probability Learning – Sample Complexity – Finite and
Infinite Hypothesis Spaces – Mistake Bound Model.

UNIT IV INSTANT BASED LEARNING 9

K- Nearest Neighbour Learning – Locally weighted Regression – Radial Basis Functions –

Case Based Learning.

UNIT V ADVANCED LEARNING 9

Learning Sets of Rules – Sequential Covering Algorithm – Learning Rule Set – First Order
Rules – Sets of First Order Rules – Induction on Inverted Deduction – Inverting Resolution –
Analytical Learning – Perfect Domain Theories – Explanation Base Learning – FOCL
Algorithm – Reinforcement Learning – Task – Q-Learning – Temporal Difference Learning

TOTAL :45 PERIODS

OUTCOMES:

At the end of the course, the students will be able to

• Differentiate between supervised, unsupervised, semi-supervised machine learning

approaches
• Discuss the decision tree algorithm and indentity and overcome the problem of
overfitting
• Discuss and apply the back propagation algorithm and genetic algorithms to various
problems
• Apply the Bayesian concepts to machine learning
• Analyse and suggest appropriate machine learning approaches for various types of
problems

TEXT BOOK:

1. Tom M. Mitchell, ―Machine Learning, McGraw-Hill Education (India) Private

Limited, 2013.

REFERENCES:

1. Ethem Alpaydin, ―Introduction to Machine Learning (Adaptive Computation and

Machine Learning), The MIT Press 2004.
2. Stephen Marsland, ―Machine Learning: An Algorithmic Perspective, CRC Press,
2009.
 1

UNIT 1
INTRODUCTION

Ever since computers were invented, we have wondered whether they might be made to learn.
If we could understand how to program them to learn-to improve automatically with
experience-the impact would be dramatic.
• Imagine computers learning from medical records which treatments are most effective
for new diseases
• Houses learning from experience to optimize energy costs based on the particular usage
patterns of their occupants.
• Personal software assistants learning the evolving interests of their users in order to
highlight especially relevant stories from the online morning newspaper

A successful understanding of how to make computers learn would open up many new uses
of computers and new levels of competence and customization

Some successful applications of machine learning

• Learning to recognize spoken words
• Learning to drive an autonomous vehicle
• Learning to classify new astronomical structures
• Learning to play world-class backgammon

Why is Machine Learning Important?

• Some tasks cannot be defined well, except by examples (e.g., recognizing people).
• Relationships and correlations can be hidden within large amounts of data. Machine
Learning/Data Mining may be able to find these relationships.
• Human designers often produce machines that do not work as well as desired in the
environments in which they are used.
• The amount of knowledge available about certain tasks might be too large for explicit
encoding by humans (e.g., medical diagnostic).
• Environments change over time.
• New knowledge about tasks is constantly being discovered by humans. It may be
difficult to continuously re-design systems “by hand”.
 2

WELL-POSED LEARNING PROBLEMS

Definition: A computer program is said to learn from experience E with respect to some class
of tasks T and performance measure P, if its performance at tasks in T, as measured by P,
improves with experience E.

To have a well-defined learning problem, three features needs to be identified:

1. The class of tasks
2. The measure of performance to be improved
3. The source of experience

Examples
1. Checkers game: A computer program that learns to play checkers might improve its
performance as measured by its ability to win at the class of tasks involving playing
checkers games, through experience obtained by playing games against itself.

Fig: Checker game board

A checkers learning problem:
• Task T: playing checkers
• Performance measure P: percent of games won against opponents
• Training experience E: playing practice games against itself

2. A handwriting recognition learning problem:

• Task T: recognizing and classifying handwritten words within images
• Performance measure P: percent of words correctly classified
• Training experience E: a database of handwritten words with given
classifications
3. A robot driving learning problem:
• Task T: driving on public four-lane highways using vision sensors
• Performance measure P: average distance travelled before an error (as judged
by human overseer)
• Training experience E: a sequence of images and steering commands recorded
while observing a human driver
 3

DESIGNING A LEARNING SYSTEM

The basic design issues and approaches to machine learning are illustrated by designing a
program to learn to play checkers, with the goal of entering it in the world checkers
tournament
1. Choosing the Training Experience
2. Choosing the Target Function
3. Choosing a Representation for the Target Function
4. Choosing a Function Approximation Algorithm
1. Estimating training values
2. Adjusting the weights
5. The Final Design

1. Choosing the Training Experience

• The first design choice is to choose the type of training experience from which the
system will learn.
• The type of training experience available can have a significant impact on success or
failure of the learner.

There are three attributes which impact on success or failure of the learner

1. Whether the training experience provides direct or indirect feedback regarding the
choices made by the performance system.

For example, in checkers game:

In learning to play checkers, the system might learn from direct training examples
consisting of individual checkers board states and the correct move for each.

Indirect training examples consisting of the move sequences and final outcomes of
various games played. The information about the correctness of specific moves early in
the game must be inferred indirectly from the fact that the game was eventually won or
lost.

Here the learner faces an additional problem of credit assignment, or determining the
degree to which each move in the sequence deserves credit or blame for the final
outcome. Credit assignment can be a particularly difficult problem because the game
can be lost even when early moves are optimal, if these are followed later by poor
moves.
Hence, learning from direct training feedback is typically easier than learning from
indirect feedback.
 4

2. The degree to which the learner controls the sequence of training examples

For example, in checkers game:

The learner might depends on the teacher to select informative board states and to
provide the correct move for each.

Alternatively, the learner might itself propose board states that it finds particularly
confusing and ask the teacher for the correct move.

The learner may have complete control over both the board states and (indirect) training
classifications, as it does when it learns by playing against itself with no teacher present.

3. How well it represents the distribution of examples over which the final system
performance P must be measured

For example, in checkers game:

In checkers learning scenario, the performance metric P is the percent of games the
system wins in the world tournament.

If its training experience E consists only of games played against itself, there is a danger
that this training experience might not be fully representative of the distribution of
situations over which it will later be tested.
It is necessary to learn from a distribution of examples that is different from those on
which the final system will be evaluated.

2. Choosing the Target Function

The next design choice is to determine exactly what type of knowledge will be learned and
how this will be used by the performance program.

Let’s consider a checkers-playing program that can generate the legal moves from any board
state.
The program needs only to learn how to choose the best move from among these legal moves.
We must learn to choose among the legal moves, the most obvious choice for the type of
information to be learned is a program, or function, that chooses the best move for any given
board state.

1. Let ChooseMove be the target function and the notation is

ChooseMove : B→ M
which indicate that this function accepts as input any board from the set of legal board
states B and produces as output some move from the set of legal moves M.
 5

ChooseMove is a choice for the target function in checkers example, but this function
will turn out to be very difficult to learn given the kind of indirect training experience
available to our system

2. An alternative target function is an evaluation function that assigns a numerical score

to any given board state
Let the target function V and the notation
V:B →R

which denote that V maps any legal board state from the set B to some real value.
Intend for this target function V to assign higher scores to better board states. If the
system can successfully learn such a target function V, then it can easily use it to select
the best move from any current board position.

Let us define the target value V(b) for an arbitrary board state b in B, as follows:
• If b is a final board state that is won, then V(b) = 100
• If b is a final board state that is lost, then V(b) = -100
• If b is a final board state that is drawn, then V(b) = 0
• If b is a not a final state in the game, then V(b) = V(b' ),

Where b' is the best final board state that can be achieved starting from b and playing optimally
until the end of the game

3. Choosing a Representation for the Target Function

Let’s choose a simple representation - for any given board state, the function c will be
calculated as a linear combination of the following board features:

• xl: the number of black pieces on the board

• x2: the number of red pieces on the board
• x3: the number of black kings on the board
• x4: the number of red kings on the board
• x5: the number of black pieces threatened by red (i.e., which can be captured on red's
next turn)
• x6: the number of red pieces threatened by black

Thus, learning program will represent as a linear function of the form

 6

Where,
• w0 through w6 are numerical coefficients, or weights, to be chosen by the learning
algorithm.
• Learned values for the weights w1 through w6 will determine the relative importance
of the various board features in determining the value of the board
• The weight w0 will provide an additive constant to the board value

4. Choosing a Function Approximation Algorithm

In order to learn the target function f we require a set of training examples, each describing a
specific board state b and the training value Vtrain(b) for b.

Each training example is an ordered pair of the form (b, Vtrain(b)).

For instance, the following training example describes a board state b in which black has won
the game (note x2 = 0 indicates that red has no remaining pieces) and for which the target
function value Vtrain(b) is therefore +100.

((x1=3, x2=0, x3=1, x4=0, x5=0, x6=0), +100)

Function Approximation Procedure

1. Derive training examples from the indirect training experience available to the learner
2. Adjusts the weights wi to best fit these training examples

1. Estimating training values

A simple approach for estimating training values for intermediate board states is to
assign the training value of Vtrain(b) for any intermediate board state b to be
V̂(Successor(b))

Where ,
• V̂is the learner's current approximation to V
• Successor(b) denotes the next board state following b for which it is again the
program's turn to move

Rule for estimating training values

Vtrain(b) ← V̂ (Successor(b))
 7

2. Adjusting the weights

Specify the learning algorithm for choosing the weights wi to best fit the set of training
examples {(b, Vtrain(b))}
A first step is to define what we mean by the bestfit to the training data.
One common approach is to define the best hypothesis, or set of weights, as that which
minimizes the squared error E between the training values and the values predicted by
the hypothesis.

Several algorithms are known for finding weights of a linear function that minimize E.
One such algorithm is called the least mean squares, or LMS training rule. For each
observed training example it adjusts the weights a small amount in the direction that
reduces the error on this training example

LMS weight update rule :- For each training example (b, Vtrain(b))
Use the current weights to calculate V̂ (b)
For each weight wi, update it as

wi ← wi + ƞ (Vtrain (b) - V̂ (b)) xi

Here ƞ is a small constant (e.g., 0.1) that moderates the size of the weight update.

Working of weight update rule

• When the error (Vtrain(b)- V̂(b)) is zero, no weights are changed.

• When (Vtrain(b) - V̂ (b)) is positive (i.e., when V̂ (b) is too low), then each weightis
increased in proportion to the value of its corresponding feature. This will raise
the value of V̂(b), reducing the error.
• If the value of some feature xi is zero, then its weight is not altered regardless of
the error, so that the only weights updated are those whose features actually occur
on the training example board.
 8

5. The Final Design

The final design of checkers learning system can be described by four distinct program modules
that represent the central components in many learning systems

1. The Performance System is the module that must solve the given performance task by
using the learned target function(s). It takes an instance of a new problem (new game)
as input and produces a trace of its solution (game history) as output.

2. The Critic takes as input the history or trace of the game and produces as output a set
of training examples of the target function

3. The Generalizer takes as input the training examples and produces an output
hypothesis that is its estimate of the target function. It generalizes from the specific
training examples, hypothesizing a general function that covers these examples and
other cases beyond the training examples.

4. The Experiment Generator takes as input the current hypothesis and outputs a new
problem (i.e., initial board state) for the Performance System to explore. Its role is to
pick new practice problems that will maximize the learning rate of the overall system.

The sequence of design choices made for the checkers program is summarized in below figure
 9

PERSPECTIVES AND ISSUES IN MACHINE LEARNING

Issues in Machine Learning

The field of machine learning, and much of this book, is concerned with answering questions
such as the following
• What algorithms exist for learning general target functions from specific training
examples? In what settings will particular algorithms converge to the desired function,
given sufficient training data? Which algorithms perform best for which types of
problems and representations?
• How much training data is sufficient? What general bounds can be found to relate the
confidence in learned hypotheses to the amount of training experience and the character
of the learner's hypothesis space?
 10

• When and how can prior knowledge held by the learner guide the process of generalizing
from examples? Can prior knowledge be helpful even when it is only approximately
correct?
• What is the best strategy for choosing a useful next training experience, and how does
the choice of this strategy alter the complexity of the learning problem?
• What is the best way to reduce the learning task to one or more function approximation
problems? Put another way, what specific functions should the system attempt to learn?
Can this process itself be automated?
• How can the learner automatically alter its representation to improve its ability to
represent and learn the target function?
 11

CONCEPT LEARNING

• Learning involves acquiring general concepts from specific training examples. Example:
People continually learn general concepts or categories such as "bird," "car," "situations in
which I should study more in order to pass the exam," etc.
• Each such concept can be viewed as describing some subset of objects or events defined
over a larger set
• Alternatively, each concept can be thought of as a Boolean-valued function defined over this
larger set. (Example: A function defined over all animals, whose value is true for birds and
false for other animals).

Definition: Concept learning - Inferring a Boolean-valued function from training examples of

its input and output

A CONCEPT LEARNING TASK

Consider the example task of learning the target concept "Days on which Aldo enjoys
his favorite water sport”

Example Sky AirTemp Humidity Wind Water Forecast EnjoySport

1 Sunny Warm Normal Strong Warm Same Yes

2 Sunny Warm High Strong Warm Same Yes

3 Rainy Cold High Strong Warm Change No

4 Sunny Warm High Strong Cool Change Yes

Table: Positive and negative training examples for the target concept EnjoySport.

The task is to learn to predict the value of EnjoySport for an arbitrary day, based on the
values of its other attributes?

What hypothesis representation is provided to the learner?

• Let’s consider a simple representation in which each hypothesis consists of a

conjunction of constraints on the instance attributes.
• Let each hypothesis be a vector of six constraints, specifying the values of the six
attributes Sky, AirTemp, Humidity, Wind, Water, and Forecast.
 12

For each attribute, the hypothesis will either

• Indicate by a "?' that any value is acceptable for this attribute,
• Specify a single required value (e.g., Warm) for the attribute, or
• Indicate by a "Φ" that no value is acceptable

If some instance x satisfies all the constraints of hypothesis h, then h classifies x as a positive
example (h(x) = 1).

The hypothesis that PERSON enjoys his favorite sport only on cold days with high humidity
is represented by the expression
(?, Cold, High, ?, ?, ?)

The most general hypothesis-that every day is a positive example-is represented by

(?, ?, ?, ?, ?, ?)

The most specific possible hypothesis-that no day is a positive example-is represented by

(Φ, Φ, Φ, Φ, Φ, Φ)

Notation

• The set of items over which the concept is defined is called the set of instances, which is
denoted by X.

Example: X is the set of all possible days, each represented by the attributes: Sky, AirTemp,
Humidity, Wind, Water, and Forecast

• The concept or function to be learned is called the target concept, which is denoted by c.
c can be any Boolean valued function defined over the instances X

c: X→ {O, 1}

Example: The target concept corresponds to the value of the attribute EnjoySport
(i.e., c(x) = 1 if EnjoySport = Yes, and c(x) = 0 if EnjoySport = No).

• Instances for which c(x) = 1 are called positive examples, or members of the target concept.
• Instances for which c(x) = 0 are called negative examples, or non-members of the target
concept.
• The ordered pair (x, c(x)) to describe the training example consisting of the instance x and
its target concept value c(x).
• D to denote the set of available training examples
 13

• The symbol H to denote the set of all possible hypotheses that the learner may consider
regarding the identity of the target concept. Each hypothesis h in H represents a Boolean-
valued function defined over X
h: X→{O, 1}

The goal of the learner is to find a hypothesis h such that h(x) = c(x) for all x in X.

• Given:
• Instances X: Possible days, each described by the attributes
• Sky (with possible values Sunny, Cloudy, and Rainy),
• AirTemp (with values Warm and Cold),
• Humidity (with values Normal and High),
• Wind (with values Strong and Weak),
• Water (with values Warm and Cool),
• Forecast (with values Same and Change).

• Hypotheses H: Each hypothesis is described by a conjunction of constraints on the

attributes Sky, AirTemp, Humidity, Wind, Water, and Forecast. The constraints may be
"?" (any value is acceptable), “Φ” (no value is acceptable), or a specific value.

• Target concept c: EnjoySport : X → {0, l}

• Training examples D: Positive and negative examples of the target function

• Determine:
• A hypothesis h in H such that h(x) = c(x) for all x in X.

Table: The EnjoySport concept learning task.

The inductive learning hypothesis

Any hypothesis found to approximate the target function well over a sufficiently large set of
training examples will also approximate the target function well over other unobserved
examples.
 14

CONCEPT LEARNING AS SEARCH

• Concept learning can be viewed as the task of searching through a large space of
hypotheses implicitly defined by the hypothesis representation.
• The goal of this search is to find the hypothesis that best fits the training examples.

Example:
Consider the instances X and hypotheses H in the EnjoySport learning task. The attribute Sky
has three possible values, and AirTemp, Humidity, Wind, Water, Forecast each have two
possible values, the instance space X contains exactly
3.2.2.2.2.2 = 96 distinct instances
5.4.4.4.4.4 = 5120 syntactically distinct hypotheses within H.

Every hypothesis containing one or more "Φ" symbols represents the empty set of instances;
that is, it classifies every instance as negative.
1 + (4.3.3.3.3.3) = 973. Semantically distinct hypotheses

General-to-Specific Ordering of Hypotheses

Consider the two hypotheses

h1 = (Sunny, ?, ?, Strong, ?, ?)
h2 = (Sunny, ?, ?, ?, ?, ?)

• Consider the sets of instances that are classified positive by hl and by h2.
• h2 imposes fewer constraints on the instance, it classifies more instances as positive. So,
any instance classified positive by hl will also be classified positive by h2. Therefore, h2
is more general than hl.

Given hypotheses hj and hk, hj is more-general-than or- equal do hk if and only if any instance
that satisfies hk also satisfies hi

Definition: Let hj and hk be Boolean-valued functions defined over X. Then hj is more general-
than-or-equal-to hk (written hj ≥ hk) if and only if

( xX ) [(hk (x) = 1) → (hj (x) = 1)]

 15

• In the figure, the box on the left represents the set X of all instances, the box on the right
the set H of all hypotheses.
• Each hypothesis corresponds to some subset of X-the subset of instances that it classifies
positive.
• The arrows connecting hypotheses represent the more - general -than relation, with the
arrow pointing toward the less general hypothesis.
• Note the subset of instances characterized by h2 subsumes the subset characterized by
hl , hence h2 is more - general– than h1

FIND-S: FINDING A MAXIMALLY SPECIFIC HYPOTHESIS

FIND-S Algorithm

1. Initialize h to the most specific hypothesis in H

2. For each positive training instance x
For each attribute constraint ai in h
If the constraint ai is satisfied by x
Then do nothing
Else replace ai in h by the next more general constraint that is satisfied by x
3. Output hypothesis h
 16

To illustrate this algorithm, assume the learner is given the sequence of training examples
from the EnjoySport task

Example Sky AirTemp Humidity Wind Water Forecast EnjoySport

1 Sunny Warm Normal Strong Warm Same Yes
2 Sunny Warm High Strong Warm Same Yes
3 Rainy Cold High Strong Warm Change No
4 Sunny Warm High Strong Cool Change Yes

• The first step of FIND-S is to initialize h to the most specific hypothesis in H

h - (Ø, Ø, Ø, Ø, Ø, Ø)

• Consider the first training example

x1 = <Sunny Warm Normal Strong Warm Same>, +

Observing the first training example, it is clear that hypothesis h is too specific. None
of the "Ø" constraints in h are satisfied by this example, so each is replaced by the next
more general constraint that fits the example
h1 = <Sunny Warm Normal Strong Warm Same>

• Consider the second training example

x2 = <Sunny, Warm, High, Strong, Warm, Same>, +

The second training example forces the algorithm to further generalize h, this time
substituting a "?" in place of any attribute value in h that is not satisfied by the new
example
h2 = <Sunny Warm ? Strong Warm Same>

• Consider the third training example

x3 = <Rainy, Cold, High, Strong, Warm, Change>, -

Upon encountering the third training the algorithm makes no change to h. The FIND-S
algorithm simply ignores every negative example.
h3 = < Sunny Warm ? Strong Warm Same>

• Consider the fourth training example

x4 = <Sunny Warm High Strong Cool Change>, +

The fourth example leads to a further generalization of h

h4 = < Sunny Warm ? Strong ? ? >
 17

The key property of the FIND-S algorithm

• FIND-S is guaranteed to output the most specific hypothesis within H that is consistent
with the positive training examples
• FIND-S algorithm’s final hypothesis will also be consistent with the negative examples
provided the correct target concept is contained in H, and provided the training examples
are correct.

Unanswered by FIND-S

1. Has the learner converged to the correct target concept?

2. Why prefer the most specific hypothesis?
3. Are the training examples consistent?
4. What if there are several maximally specific consistent hypotheses?
 18

VERSION SPACES AND THE CANDIDATE-ELIMINATION ALGORITHM

The key idea in the CANDIDATE-ELIMINATION algorithm is to output a description of the

set of all hypotheses consistent with the training examples

Representation

Definition: consistent- A hypothesis h is consistent with a set of training examples D if and

only if h(x) = c(x) for each example (x, c(x)) in D.

Consistent (h, D)  ( x, c(x)  D) h(x) = c(x))

Note difference between definitions of consistent and satisfies

• An example x is said to satisfy hypothesis h when h(x) = 1, regardless of whether x is
a positive or negative example of the target concept.
• An example x is said to consistent with hypothesis h iff h(x) = c(x)

Definition: version space- The version space, denoted V S with respect to hypothesis space
H, D
H and training examples D, is the subset of hypotheses from H consistent with the training
examples in D
V S {h  H | Consistent (h, D)}
H, D

The LIST-THEN-ELIMINATION algorithm

The LIST-THEN-ELIMINATE algorithm first initializes the version space to contain all
hypotheses in H and then eliminates any hypothesis found inconsistent with any training
example.

1. VersionSpace c a list containing every hypothesis in H

2. For each training example, (x, c(x))
remove from VersionSpace any hypothesis h for which h(x) ≠ c(x)
3. Output the list of hypotheses in VersionSpace

The LIST-THEN-ELIMINATE Algorithm

• List-Then-Eliminate works in principle, so long as version space is finite.

• However, since it requires exhaustive enumeration of all hypotheses in practice it is not
feasible.
 19

A More Compact Representation for Version Spaces

The version space is represented by its most general and least general members. These
members form general and specific boundary sets that delimit the version space within the
partially ordered hypothesis space.

Definition: The general boundary G, with respect to hypothesis space H and training data D,
is the set of maximally general members of H consistent with D

G {g  H | Consistent (g, D)(g'  H)[(g'  g)  Consistent(g', D)]}

g

Definition: The specific boundary S, with respect to hypothesis space H and training data D,
is the set of minimally general (i.e., maximally specific) members of H consistent with D.

S {s  H | Consistent (s, D)(s'  H)[(s  s')  Consistent(s', D)]}

g

Theorem: Version Space representation theorem

Theorem: Let X be an arbitrary set of instances and Let H be a set of Boolean-valued

hypotheses defined over X. Let c: X →{O, 1} be an arbitrary target concept defined over X,
and let D be an arbitrary set of training examples {(x, c(x))). For all X, H, c, and D such that S
and G are well defined,

VS ={ h  H | (s  S ) (g  G ) ( g  h  s )}
H,D g g

To Prove:
1. Every h satisfying the right hand side of the above expression is in VS
H, D
2. Every member of VS satisfies the right-hand side of the expression
H, D

Sketch of proof:
1. let g, h, s be arbitrary members of G, H, S respectively with g g h g s
• By the definition of S, s must be satisfied by all positive examples in D. Because h g s,
h must also be satisfied by all positive examples in D.
• By the definition of G, g cannot be satisfied by any negative example in D, and because
g g h h cannot be satisfied by any negative example in D. Because h is satisfied by all
positive examples in D and by no negative examples in D, h is consistent with D, and
therefore h is a member of VSH,D.
2. It can be proven by assuming some h in VSH,D,that does not satisfy the right-hand side
of the expression, then showing that this leads to an inconsistency
 20

CANDIDATE-ELIMINATION Learning Algorithm

The CANDIDATE-ELIMINTION algorithm computes the version space containing all

hypotheses from H that are consistent with an observed sequence of training examples.

Initialize G to the set of maximally general hypotheses in H

Initialize S to the set of maximally specific hypotheses in H
For each training example d, do
• If d is a positive example
• Remove from G any hypothesis inconsistent with d
• For each hypothesis s in S that is not consistent with d
• Remove s from S
• Add to S all minimal generalizations h of s such that
• h is consistent with d, and some member of G is more general than h
• Remove from S any hypothesis that is more general than another hypothesis in S

• If d is a negative example
• Remove from S any hypothesis inconsistent with d
• For each hypothesis g in G that is not consistent with d
• Remove g from G
• Add to G all minimal specializations h of g such that
• h is consistent with d, and some member of S is more specific than h
• Remove from G any hypothesis that is less general than another hypothesis in G

CANDIDATE- ELIMINTION algorithm using version spaces

An Illustrative Example

Example Sky AirTemp Humidity Wind Water Forecast EnjoySport

1 Sunny Warm Normal Strong Warm Same Yes
2 Sunny Warm High Strong Warm Same Yes
3 Rainy Cold High Strong Warm Change No
4 Sunny Warm High Strong Cool Change Yes
 21

CANDIDATE-ELIMINTION algorithm begins by initializing the version space to the

set of all hypotheses in H;

Initializing the G boundary set to contain the most general hypothesis in H

G0 ?, ?, ?, ?, ?, ?

Initializing the S boundary set to contain the most specific (least general) hypothesis
S0 , , , , , 

• When the first training example is presented, the CANDIDATE-ELIMINTION algorithm

checks the S boundary and finds that it is overly specific and it fails to cover the positive
example.
• The boundary is therefore revised by moving it to the least more general hypothesis that
covers this new example
• No update of the G boundary is needed in response to this training example because Go
correctly covers this example

• When the second training example is observed, it has a similar effect of generalizing S
further to S2, leaving G again unchanged i.e., G2 = G1 = G0
 22

• Consider the third training example. This negative example reveals that the G boundary
of the version space is overly general, that is, the hypothesis in G incorrectly predicts
that this new example is a positive example.
• The hypothesis in the G boundary must therefore be specialized until it correctly
classifies this new negative example

Given that there are six attributes that could be specified to specialize G 2, why are there only
three new hypotheses in G3?
For example, the hypothesis h = (?, ?, Normal, ?, ?, ?) is a minimal specialization of G2
that correctly labels the new example as a negative example, but it is not included in G3.
The reason this hypothesis is excluded is that it is inconsistent with the previously
encountered positive examples

• Consider the fourth training example.

 23

• This positive example further generalizes the S boundary of the version space. It also
results in removing one member of the G boundary, because this member fails to
cover the new positive example

After processing these four examples, the boundary sets S4 and G4 delimit the version space
of all hypotheses consistent with the set of incrementally observed training examples.
 24

INDUCTIVE BIAS

The fundamental questions for inductive inference

1. What if the target concept is not contained in the hypothesis space?

2. Can we avoid this difficulty by using a hypothesis space that includes every possible
hypothesis?
3. How does the size of this hypothesis space influence the ability of the algorithm to
generalize to unobserved instances?
4. How does the size of the hypothesis space influence the number of training examples
that must be observed?

These fundamental questions are examined in the context of the CANDIDATE-

ELIMINTION algorithm

A Biased Hypothesis Space

• Suppose the target concept is not contained in the hypothesis space H, then obvious
solution is to enrich the hypothesis space to include every possible hypothesis.
• Consider the EnjoySport example in which the hypothesis space is restricted to include
only conjunctions of attribute values. Because of this restriction, the hypothesis space is
unable to represent even simple disjunctive target concepts such as
"Sky = Sunny or Sky = Cloudy."
• The following three training examples of disjunctive hypothesis, the algorithm would
find that there are zero hypotheses in the version space

Sunny Warm Normal Strong Cool Change Y

Cloudy Warm Normal Strong Cool Change Y
Rainy Warm Normal Strong Cool Change N

• If Candidate Elimination algorithm is applied, then it end up with empty Version Space.
After first two training example
S= ? Warm Normal Strong Cool Change

• This new hypothesis is overly general and it incorrectly covers the third negative
training example! So H does not include the appropriate c.
• In this case, a more expressive hypothesis space is required.
 25

An Unbiased Learner

• The solution to the problem of assuring that the target concept is in the hypothesis space H
is to provide a hypothesis space capable of representing every teachable concept that is
representing every possible subset of the instances X.
• The set of all subsets of a set X is called the power set of X

• In the EnjoySport learning task the size of the instance space X of days described by
the six attributes is 96 instances.
• Thus, there are 296 distinct target concepts that could be defined over this instance space
and learner might be called upon to learn.
• The conjunctive hypothesis space is able to represent only 973 of these - a biased
hypothesis space indeed

• Let us reformulate the EnjoySport learning task in an unbiased way by defining a new
hypothesis space H' that can represent every subset of instances
• The target concept "Sky = Sunny or Sky = Cloudy" could then be described as

(Sunny, ?, ?, ?, ?, ?) v (Cloudy, ?, ?, ?, ?, ?)

The Futility of Bias-Free Learning

Inductive learning requires some form of prior assumptions, or inductive bias

Definition:
Consider a concept learning algorithm L for the set of instances X.
• Let c be an arbitrary concept defined over X
• Let D = {(x , c(x))} be an arbitrary set of training examples of c.
c
• Let L (x , D ) denote the classification assigned to the instance x by L after training on
i c i
the data D .
c
• The inductive bias of L is any minimal set of assertions B such that for any target concept
c and corresponding training examples D
c

• ( xi  X ) [(B  Dc  xi) ├ L (xi, Dc )]

 26

The below figure explains

• Modelling inductive systems by equivalent deductive systems.
• The input-output behavior of the CANDIDATE-ELIMINATION algorithm using a
hypothesis space H is identical to that of a deductive theorem prover utilizing the
assertion "H contains the target concept." This assertion is therefore called the inductive
bias of the CANDIDATE-ELIMINATION algorithm.
• Characterizing inductive systems by their inductive bias allows modelling them by their
equivalent deductive systems. This provides a way to compare inductive systems
according to their policies for generalizing beyond the observed training data.
 27

DECISION TREE LEARNING

Decision tree learning is a method for approximating discrete-valued target functions, in which
the learned function is represented by a decision tree.

DECISION TREE REPRESENTATION

• Decision trees classify instances by sorting them down the tree from the root to some
leaf node, which provides the classification of the instance.
• Each node in the tree specifies a test of some attribute of the instance, and each branch
descending from that node corresponds to one of the possible values for this attribute.
• An instance is classified by starting at the root node of the tree, testing the attribute
specified by this node, then moving down the tree branch corresponding to the value of
the attribute in the given example. This process is then repeated for the subtree rooted
at the new node.

FIGURE: A decision tree for the concept PlayTennis. An example is classified by sorting it
through the tree to the appropriate leaf node, then returning the classification associated with
this leaf
 28

• Decision trees represent a disjunction of conjunctions of constraints on the attribute

values of instances.
• Each path from the tree root to a leaf corresponds to a conjunction of attribute tests,
and the tree itself to a disjunction of these conjunctions

For example, the decision tree shown in above figure corresponds to the expression
(Outlook = Sunny ∧ Humidity = Normal)
∨ (Outlook = Overcast)
∨ (Outlook = Rain ∧ Wind = Weak)

APPROPRIATE PROBLEMS FOR DECISION TREE

LEARNING
Decision tree learning is generally best suited to problems with the following characteristics:

1. Instances are represented by attribute-value pairs – Instances are described by a

fixed set of attributes and their values

2. The target function has discrete output values – The decision tree assigns a Boolean
classification (e.g., yes or no) to each example. Decision tree methods easily extend to
learning functions with more than two possible output values.

3. Disjunctive descriptions may be required

4. The training data may contain errors – Decision tree learning methods are robust to
errors, both errors in classifications of the training examples and errors in the attribute
values that describe these examples.

5. The training data may contain missing attribute values – Decision tree methods can
be used even when some training examples have unknown values
 29

THE BASIC DECISION TREE LEARNING ALGORITHM

The basic algorithm is ID3 which learns decision trees by constructing them top-down

ID3(Examples, Target_attribute, Attributes)

Examples are the training examples. Target_attribute is the attribute whose value is to be
predicted by the tree. Attributes is a list of other attributes that may be tested by the
learned decision tree. Returns a decision tree that correctly classifies the given Examples.

• Create a Root node for the tree

• If all Examples are positive, Return the single-node tree Root, with label = +
• If all Examples are negative, Return the single-node tree Root, with label = -
• If Attributes is empty, Return the single-node tree Root, with label = most common value
of Target_attribute in Examples

• Otherwise Begin
• A ← the attribute from Attributes that best* classifies Examples
• The decision attribute for Root ← A
• For each possible value, vi, of A,
• Add a new tree branch below Root, corresponding to the test A = vi
• Let Examples vi, be the subset of Examples that have value vi for A
• If Examples vi , is empty
• Then below this new branch add a leaf node with label = most common
value of Target_attribute in Examples
• Else below this new branch add the subtree
ID3(Examples vi, Targe_tattribute, Attributes – {A}))
• End
• Return Root

* The best attribute is the one with highest information gain

TABLE: Summary of the ID3 algorithm specialized to learning Boolean-valued functions. ID3
is a greedy algorithm that grows the tree top-down, at each node selecting the attribute that best
classifies the local training examples. This process continues until the tree perfectly classifies
the training examples, or until all attributes have been used
 30

Which Attribute Is the Best Classifier?

• The central choice in the ID3 algorithm is selecting which attribute to test at each node
in the tree.
• A statistical property called information gain that measures how well a given attribute
separates the training examples according to their target classification.
• ID3 uses information gain measure to select among the candidate attributes at each
step while growing the tree.

ENTROPY MEASURES HOMOGENEITY OF EXAMPLES

To define information gain, we begin by defining a measure called entropy. Entropy

measures the impurity of a collection of examples.

Given a collection S, containing positive and negative examples of some target concept, the
entropy of S relative to this Boolean classification is

Where,
p+ is the proportion of positive examples in S
p- is the proportion of negative examples in S.

Example:
Suppose S is a collection of 14 examples of some boolean concept, including 9 positive and 5
negative examples. Then the entropy of S relative to this boolean classification is

• The entropy is 0 if all members of S belong to the same class

• The entropy is 1 when the collection contains an equal number of positive and negative
examples
• If the collection contains unequal numbers of positive and negative examples, the
entropy is between 0 and 1
 31

INFORMATION GAIN MEASURES THE EXPECTED REDUCTION IN ENTROPY

• Information gain, is the expected reduction in entropy caused by partitioning the

examples according to this attribute.
• The information gain, Gain(S, A) of an attribute A, relative to a collection of examples
S, is defined as

Example: Information gain

Let, Values(Wind) = {Weak, Strong}

S = [9+, 5−]
S = [6+, 2−]
Weak
S = [3+, 3−]
Strong

Information gain of attribute Wind:

Gain(S, Wind) = Entropy(S) − 8/14 Entropy (SWeak) − 6/14 Entropy (SStrong)
= 0.94 – (8/14)* 0.811 – (6/14) *1.00
= 0.048
 32

An Illustrative Example

• To illustrate the operation of ID3, consider the learning task represented by the training
examples of below table.
• Here the target attribute PlayTennis, which can have values yes or no for different days.
• Consider the first step through the algorithm, in which the topmost node of the decision
tree is created.

Day Outlook Temperature Humidity Wind PlayTennis

D1 Sunny Hot High Weak No
D2 Sunny Hot High Strong No
D3 Overcast Hot High Weak Yes
D4 Rain Mild High Weak Yes
D5 Rain Cool Normal Weak Yes
D6 Rain Cool Normal Strong No
D7 Overcast Cool Normal Strong Yes
D8 Sunny Mild High Weak No
D9 Sunny Cool Normal Weak Yes
D10 Rain Mild Normal Weak Yes
D11 Sunny Mild Normal Strong Yes
D12 Overcast Mild High Strong Yes
D13 Overcast Hot Normal Weak Yes
D14 Rain Mild High Strong No

• ID3 determines the information gain for each candidate attribute (i.e., Outlook,
Temperature, Humidity, and Wind), then selects the one with highest information gain.
 33

• The information gain values for all four attributes are

Gain(S, Outlook) = 0.246
Gain(S, Humidity) = 0.151
Gain(S, Wind) = 0.048
Gain(S, Temperature) = 0.029

• According to the information gain measure, the Outlook attribute provides the best
prediction of the target attribute, PlayTennis, over the training examples. Therefore,
Outlook is selected as the decision attribute for the root node, and branches are created
below the root for each of its possible values i.e., Sunny, Overcast, and Rain.
 34

SRain = { D4, D5, D6, D10, D14}

Gain (SRain , Humidity) = 0.970 – (2/5)1.0 – (3/5)0.917 = 0.019

Gain (SRain , Temperature) =0.970 – (0/5)0.0 – (3/5)0.918 – (2/5)1.0 = 0.019
Gain (SRain , Wind) =0.970 – (3/5)0.0 – (2/5)0.0 = 0.970
 35

HYPOTHESIS SPACE SEARCH IN DECISION TREE

LEARNING
• ID3 can be characterized as searching a space of hypotheses for one that fits the training
examples.
• The hypothesis space searched by ID3 is the set of possible decision trees.
• ID3 performs a simple-to complex, hill-climbing search through this hypothesis space,
beginning with the empty tree, then considering progressively more elaborate
hypotheses in search of a decision tree that correctly classifies the training data

Figure: Hypothesis space search by ID3. ID3 searches through the space of possible decision
trees from simplest to increasingly complex, guided by the information gain heuristic.

By viewing ID3 in terms of its search space and search strategy, there are some insight into its
capabilities and limitations

1. ID3's hypothesis space of all decision trees is a complete space of finite discrete-valued
functions, relative to the available attributes. Because every finite discrete-valued
function can be represented by some decision tree
ID3 avoids one of the major risks of methods that search incomplete hypothesis spaces:
that the hypothesis space might not contain the target function.
 36

2. ID3 maintains only a single current hypothesis as it searches through the space of
decision trees.
For example, with the earlier version space candidate elimination method, which
maintains the set of all hypotheses consistent with the available training examples.

By determining only a single hypothesis, ID3 loses the capabilities that follow from
explicitly representing all consistent hypotheses.
For example, it does not have the ability to determine how many alternative decision
trees are consistent with the available training data, or to pose new instance queries that
optimally resolve among these competing hypotheses

3. ID3 in its pure form performs no backtracking in its search. Once it selects an attribute
to test at a particular level in the tree, it never backtracks to reconsider this choice.
In the case of ID3, a locally optimal solution corresponds to the decision tree it selects
along the single search path it explores. However, this locally optimal solution may be
less desirable than trees that would have been encountered along a different branch of
the search.

4. ID3 uses all training examples at each step in the search to make statistically based
decisions regarding how to refine its current hypothesis.
One advantage of using statistical properties of all the examples is that the resulting
search is much less sensitive to errors in individual training examples.
ID3 can be easily extended to handle noisy training data by modifying its termination
criterion to accept hypotheses that imperfectly fit the training data.

INDUCTIVE BIAS IN DECISION TREE LEARNING

Inductive bias is the set of assumptions that, together with the training data, deductively justify
the classifications assigned by the learner to future instances

Given a collection of training examples, there are typically many decision trees consistent with
these examples. Which of these decision trees does ID3 choose?

ID3 search strategy

• Selects in favour of shorter trees over longer ones
• Selects trees that place the attributes with highest information gain closest to the root.
 37

Approximate inductive bias of ID3: Shorter trees are preferred over larger trees

• Consider an algorithm that begins with the empty tree and searches breadth first through
progressively more complex trees.
• First considering all trees of depth 1, then all trees of depth 2, etc.
• Once it finds a decision tree consistent with the training data, it returns the smallest
consistent tree at that search depth (e.g., the tree with the fewest nodes).
• Let us call this breadth-first search algorithm BFS-ID3.
• BFS-ID3 finds a shortest decision tree and thus exhibits the bias "shorter trees are
preferred over longer trees.

A closer approximation to the inductive bias of ID3: Shorter trees are preferred over longer
trees. Trees that place high information gain attributes close to the root are preferred over
those that do not.

• ID3 can be viewed as an efficient approximation to BFS-ID3, using a greedy heuristic

search to attempt to find the shortest tree without conducting the entire breadth-first
search through the hypothesis space.
• Because ID3 uses the information gain heuristic and a hill climbing strategy, it exhibits
a more complex bias than BFS-ID3.
• In particular, it does not always find the shortest consistent tree, and it is biased to favour
trees that place attributes with high information gain closest to the root.

Restriction Biases and Preference Biases

Difference between the types of inductive bias exhibited by ID3 and by the CANDIDATE-
ELIMINATION Algorithm.
ID3:
• ID3 searches a complete hypothesis space
• It searches incompletely through this space, from simple to complex hypotheses, until
its termination condition is met
• Its inductive bias is solely a consequence of the ordering of hypotheses by its search
strategy. Its hypothesis space introduces no additional bias

CANDIDATE-ELIMINATION Algorithm:
• The version space CANDIDATE-ELIMINATION Algorithm searches an incomplete
hypothesis space
• It searches this space completely, finding every hypothesis consistent with the training
data.
• Its inductive bias is solely a consequence of the expressive power of its hypothesis
representation. Its search strategy introduces no additional bias
 38

Preference bias – The inductive bias of ID3 is a preference for certain hypotheses over others
(e.g., preference for shorter hypotheses over larger hypotheses), with no hard restriction on the
hypotheses that can be eventually enumerated. This form of bias is called a preference bias or
a search bias.

Restriction bias – The bias of the CANDIDATE ELIMINATION algorithm is in the form of a
categorical restriction on the set of hypotheses considered. This form of bias is typically called
a restriction bias or a language bias.

Which type of inductive bias is preferred in order to generalize beyond the training data, a
preference bias or restriction bias?

• A preference bias is more desirable than a restriction bias, because it allows the learner
to work within a complete hypothesis space that is assured to contain the unknown target
function.
• In contrast, a restriction bias that strictly limits the set of potential hypotheses is
generally less desirable, because it introduces the possibility of excluding the unknown
target function altogether.

Why Prefer Short

Hypotheses? Occam's razor

• Occam's razor: is the problem-solving principle that the simplest solution tends to be
the right one. When presented with competing hypotheses to solve a problem, one
should select the solution with the fewest assumptions.

• Occam's razor: “Prefer the simplest hypothesis that fits the data”.

Argument in favour of Occam’s razor:

• Fewer short hypotheses than long ones:

• Short hypotheses fits the training data which are less likely to be coincident
• Longer hypotheses fits the training data might be coincident.

• Many complex hypotheses that fit the current training data but fail to generalize
correctly to subsequent data.
 39

Argument opposed:
• There are few small trees, and our priori chance of finding one consistent with an
arbitrary set of data is therefore small. The difficulty here is that there are very many
small sets of hypotheses that one can define but understood by fewer learner.
• The size of a hypothesis is determined by the representation used internally by the
learner. Occam's razor will produce two different hypotheses from the same training
examples when it is applied by two learners, both justifying their contradictory
conclusions by Occam's razor. On this basis we might be tempted to reject Occam's
razor altogether.

ISSUES IN DECISION TREE LEARNING

Issues in learning decision trees include
1. Avoiding Overfitting the Data
Reduced error pruning
Rule post-pruning
2. Incorporating Continuous-Valued Attributes
3. Alternative Measures for Selecting Attributes
4. Handling Training Examples with Missing Attribute Values
5. Handling Attributes with Differing Costs

1. Avoiding Overfitting the Data

• The ID3 algorithm grows each branch of the tree just deeply enough to perfectly classify
the training examples but it can lead to difficulties when there is noise in the data, or
when the number of training examples is too small to produce a representative sample
of the true target function. This algorithm can produce trees that overfit the training
examples.

• Definition - Overfit: Given a hypothesis space H, a hypothesis h ∈ H is said to overfit

the training data if there exists some alternative hypothesis h' ∈ H, such that h has
smaller error than h' over the training examples, but h' has a smaller error than h over
the entire distribution of instances.
 40

The below figure illustrates the impact of overfitting in a typical application of decision tree
learning.

• The horizontal axis of this plot indicates the total number of nodes in the decision tree,
as the tree is being constructed. The vertical axis indicates the accuracy of predictions
made by the tree.
• The solid line shows the accuracy of the decision tree over the training examples. The
broken line shows accuracy measured over an independent set of test example
• The accuracy of the tree over the training examples increases monotonically as the tree
is grown. The accuracy measured over the independent test examples first increases,
then decreases.

How can it be possible for tree h to fit the training examples better than h', but for it to perform
more poorly over subsequent examples?
1. Overfitting can occur when the training examples contain random errors or noise
2. When small numbers of examples are associated with leaf nodes.

Noisy Training Example

• Example 15: <Sunny, Hot, Normal, Strong, ->

• Example is noisy because the correct label is +
• Previously constructed tree misclassifies it
 41

Approaches to avoiding overfitting in decision tree learning

• Pre-pruning (avoidance): Stop growing the tree earlier, before it reaches the point where
it perfectly classifies the training data
• Post-pruning (recovery): Allow the tree to overfit the data, and then post-prune the tree

Criterion used to determine the correct final tree size

• Use a separate set of examples, distinct from the training examples, to evaluate the utility
of post-pruning nodes from the tree
• Use all the available data for training, but apply a statistical test to estimate whether
expanding (or pruning) a particular node is likely to produce an improvement beyond
the training set
• Use measure of the complexity for encoding the training examples and the decision tree,
halting growth of the tree when this encoding size is minimized. This approach is called
the Minimum Description Length

MDL – Minimize : size(tree) + size (misclassifications(tree))

 42

Reduced-Error Pruning

• Reduced-error pruning, is to consider each of the decision nodes in the tree to be

candidates for pruning
• Pruning a decision node consists of removing the subtree rooted at that node, making
it a leaf node, and assigning it the most common classification of the training examples
affiliated with that node
• Nodes are removed only if the resulting pruned tree performs no worse than-the original
over the validation set.
• Reduced error pruning has the effect that any leaf node added due to coincidental
regularities in the training set is likely to be pruned because these same coincidences are
unlikely to occur in the validation set

The impact of reduced-error pruning on the accuracy of the decision tree is illustrated in below
figure

• The additional line in figure shows accuracy over the test examples as the tree is pruned.
When pruning begins, the tree is at its maximum size and lowest accuracy over the test
set. As pruning proceeds, the number of nodes is reduced and accuracy over the test set
increases.
• The available data has been split into three subsets: the training examples, the validation
examples used for pruning the tree, and a set of test examples used to provide an
unbiased estimate of accuracy over future unseen examples. The plot shows accuracy
over the training and test sets.
 43

Pros and Cons

Pro: Produces smallest version of most accurate T (subtree of T)
Con: Uses less data to construct T
Can afford to hold out Dvalidation?. If not (data is too limited), may make error worse
(insufficient D )
train

Rule Post-Pruning
Rule post-pruning is successful method for finding high accuracy hypotheses

• Rule post-pruning involves the following steps:

• Infer the decision tree from the training set, growing the tree until the training data is fit
as well as possible and allowing overfitting to occur.
• Convert the learned tree into an equivalent set of rules by creating one rule for each path
from the root node to a leaf node.
• Prune (generalize) each rule by removing any preconditions that result in improving its
estimated accuracy.
• Sort the pruned rules by their estimated accuracy, and consider them in this sequence
when classifying subsequent instances.

Converting a Decision Tree into Rules

 44

For example, consider the decision tree. The leftmost path of the tree in below figure is
translated into the rule.
IF (Outlook = Sunny) ^ (Humidity = High)
THEN PlayTennis = No

Given the above rule, rule post-pruning would consider removing the preconditions
(Outlook = Sunny) and (Humidity = High)

• It would select whichever of these pruning steps produced the greatest improvement in
estimated rule accuracy, then consider pruning the second precondition as a further
pruning step.
• No pruning step is performed if it reduces the estimated rule accuracy.

There are three main advantages by converting the decision tree to rules before pruning

1. Converting to rules allows distinguishing among the different contexts in which a

decision node is used. Because each distinct path through the decision tree node
produces a distinct rule, the pruning decision regarding that attribute test can be made
differently for each path.
2. Converting to rules removes the distinction between attribute tests that occur near the
root of the tree and those that occur near the leaves. Thus, it avoid messy bookkeeping
issues such as how to reorganize the tree if the root node is pruned while retaining part
of the subtree below this test.
3. Converting to rules improves readability. Rules are often easier for to understand.

2. Incorporating Continuous-Valued Attributes

Continuous-valued decision attributes can be incorporated into the learned tree.

There are two methods for Handling Continuous Attributes

1. Define new discrete valued attributes that partition the continuous attribute value into a
discrete set of intervals.
E.g., {high ≡ Temp > 35º C, med ≡ 10º C < Temp ≤ 35º C, low ≡ Temp ≤ 10º C}

2. Using thresholds for splitting nodes

e.g., A ≤ a produces subsets A ≤ a and A > a
 45

What threshold-based Boolean attribute should be defined based on Temperature?

• Pick a threshold, c, that produces the greatest information gain

• In the current example, there are two candidate thresholds, corresponding to the values
of Temperature at which the value of PlayTennis changes: (48 + 60)/2, and (80 + 90)/2.
• The information gain can then be computed for each of the candidate attributes,
Temperature >54, and Temperature >85 and the best can be selected (Temperature >54)

3. Alternative Measures for Selecting Attributes

• The problem is if attributes with many values, Gain will select it ?

• Example: consider the attribute Date, which has a very large number of possible values.
(e.g., March 4, 1979).
• If this attribute is added to the PlayTennis data, it would have the highest information
gain of any of the attributes. This is because Date alone perfectly predicts the target
attribute over the training data. Thus, it would be selected as the decision attribute for
the root node of the tree and lead to a tree of depth one, which perfectly classifies the
training data.
• This decision tree with root node Date is not a useful predictor because it perfectly
separates the training data, but poorly predict on subsequent examples.

One Approach: Use GainRatio instead of Gain

The gain ratio measure penalizes attributes by incorporating a split information, that is sensitive
to how broadly and uniformly the attribute splits the data

Where, Si is subset of S, for which attribute A has value vi

 46

4. Handling Training Examples with Missing Attribute Values

The data which is available may contain missing values for some
attributes Example: Medical diagnosis
• <Fever = true, Blood-Pressure = normal, …, Blood-Test = ?, …>
• Sometimes values truly unknown, sometimes low priority (or cost too high)

Strategies for dealing with the missing attribute value

• If node n test A, assign most common value of A among other training examples
sorted to node n
• Assign most common value of A among other training examples with same target value
• Assign a probability pi to each of the possible values vi of A rather than simply
assigning the most common value to A(x)

5. Handling Attributes with Differing Costs

• In some learning tasks the instance attributes may have associated costs.
• For example: In learning to classify medical diseases, the patients described in terms
of attributes such as Temperature, BiopsyResult, Pulse, BloodTestResults, etc.
• These attributes vary significantly in their costs, both in terms of monetary cost and
cost to patient comfort
• Decision trees use low-cost attributes where possible, depends only on high-cost
attributes only when needed to produce reliable classifications

How to Learn A Consistent Tree with Low Expected Cost?

One approach is replace Gain by Cost-Normalized-Gain

Examples of normalization functions

 47

INDUCTIVE BIAS

Why Inductive Bias?

In the Candidate-Elimination Algorithm, we get two hypotheses, one specific and one general at
the end as a final solution. Now, we also need to check if the hypothesis we got from the
algorithm is actually correct or not, also make decisions like what training examples should the
machine learn next.

Some of the fundamental questions for inductive reference are,

• What happens if the target concept isn’t in the hypothesis space?

• Will the Candidate elimination algorithm give us correct hypothesis?
• What training examples should the learners request next?

what is Inductive learning?

• This basically means learning from examples, learning on the go. [From examples we
derive rules.]
• We are given input samples (x) and output samples (f(x)) in the context of inductive
learning, and the objective is to estimate the function (f).
• The goal is to generalize [general procedure] from the samples and map such that the
output may be estimated for fresh samples in the future.

what is Deductive learning?

• Learners are initially exposed to concepts and generalizations, followed by particular

examples and exercises to aid learning.
• Already existing rules are applied to the training examples.

Biased Hypothesis Space

• It does not include all types of training instances. [ does not consider all types of training
examples.]
• The issue is that we have skewed the learner’s thinking to only evaluate conjunctive
possibilities.
• In this instance, a more expressive hypothesis space is required. [solution is, include all
hypothesis]
• Example: sunny warm normal strong cool change= yes

Unbiased Hypothesis Space:

 48

• Providing a hypothesis capable of representing set of all example.

• Possible instances:
• In the EnjoySport learning task the size of the instance space X of days described by the
six attributes in 96 instances.
• Target concept (examples): 2power96 (huge value, it is practically not possible.)

What is inductive bias?

• The inductive bias (also known as learning bias) of a learning algorithm is the set of
assumptions that the learner uses to predict outputs of given inputs that it has not
encountered.
• Inductive bias is anything which makes the algorithm learn one pattern instead of
another pattern.
• In machine learning, one aims to construct algorithms that are able to learn to predict a
certain target output.
• The idea of inductive bias is to let the learner generalize beyond the observed training
examples to infer new examples.
• ‘ > ’ -> Inductively inferred from.
• For example,
• x > y means y is inductively deduced from x.
Example
Learning algorithm – L (based on the learning algorithm you got a training data)
Training data D = {x, c(x)}
New instance = xi
Represented as L(xi, D)
(D^Xi)>L(Xi,D)

Types of Inductive Bias in ML

Maximum conditional independence

• It aims to maximize conditional independence if the hypothesis can be framed within a
 49

Bayesian framework. The Naive Bayes classifier employs this bias.

Minimum cross-validation error:
• It picks the hypothesis with the lowest cross-validation error when trying to decide
between them.
• Despite the fact that cross-validation may appear to be bias-free, the "no free lunch"
theorems demonstrate that cross-validation is in fact biased.
Maximum margin:
• When dividing a group of students into two classes, try to make the boundary as wide as
possible. The bias in support vector machines is this. It is assumed that different classes
often have a lot of space between them.
Minimum description length:
• When formulating a hypothesis, make an effort to keep the description as brief as
possible.
Minimum features:
• Unless a feature is supported by solid evidence, it should be removed. The underlying
premise of feature selection algorithms is this.
Nearest neighbors:
• In a small neighborhood in feature space, it is reasonable to assume that the majority of
the cases belong to the same class. The k-nearest neighbors' algorithm employs this bias.
• The underlying premise is that cases that are close to one another typically belong to the
same class.

HEURISTIC SPACE SEARCH

Heuristic space search

• A heuristic is a technique that is used to solve a problem faster than the classic methods.
• These techniques are used to find the approximate solution of a problem.
• Heuristics are said to be the problem-solving techniques that result in practical and quick
solutions.
• Heuristic search is class of method which is used in order to search a solution space for an
optimal solution for a problem.
• It is often possible to model the process of solving a problem as a search through a
solution space starting from an initial possible solution, with a method or set of rules
dictating how to move from one possible solution to another.
• This method or rule set must be applied repeatedly in order to eventually satisfy some
goal condition which indicates that a solution has been found.

Why do we need heuristics?

• Heuristics are used in situations in which there is the requirement of a short-term solution.
• On facing complex situations with limited resources and time, Heuristics can help to
make quick decisions by shortcuts and approximated calculations.
• Most of the heuristic methods involve mental shortcuts to make decisions on past
experiences.
 50

Common Uses for Heuristic

• Reduce mental effort needed to make decisions.

• Simplify complex and difficulty questions.
• They are a fast and accurate way to arrive at a conclusion.
• Help with problem solving.
The heuristic method might not always provide us the finest solution, but it is assured that it
helps us find a good solution in a reasonable time.

Heuristic search techniques

• Hill Climbing
• Constraint Satisfaction Problems
• Simulated Annealing
• Best First search (BFS)
We can perform the Heuristic techniques into two categories:

• Direct Heuristic Search techniques

• Weak Heuristic Search techniques
Direct Heuristic Search techniques

• It includes Blind Search, Uninformed Search, and Blind control strategy.

• These search techniques are not always possible as they require much memory and time.
• These techniques search the complete space for a solution and use the arbitrary ordering
of operations.
• The examples of Direct Heuristic search techniques include Breadth-First Search (BFS)
and Depth First Search (DFS).
Weak Heuristic Search techniques
• It includes Informed Search, Heuristic Search, and Heuristic control strategy.
• These techniques are helpful when they are applied properly to the right types of tasks.
They usually require domain-specific information.
• The examples of Weak Heuristic search techniques include Best First Search (BFS) and
A* search.
__________________________________
 UNIT 2
NEURAL NETWORKS AND GENETIC
ALGORITHMS

INTRODUCTION

Artificial neural networks (ANNs) provide a general, practical method for learning real-valued,
discrete-valued, and vector-valued target functions.

Biological Motivation

• The study of artificial neural networks (ANNs) has been inspired by the observation that
biological learning systems are built of very complex webs of interconnected Neurons
• Human information processing system consists of brain neuron: basic building block
cell that communicates information to and from various parts of body

Facts of Human Neurobiology

• Number of neurons ~ 1011

• Connection per neuron ~ 10 4 – 5
• Neuron switching time ~ 0.001 second or 10 -3
• Scene recognition time ~ 0.1 second
• 100 inference steps doesn’t seem like enough
• Highly parallel computation based on distributed representation

Properties of Neural Networks

• Many neuron-like threshold switching units

• Many weighted interconnections among units
• Highly parallel, distributed process
• Emphasis on tuning weights automatically
• Input is a high-dimensional discrete or real-valued (e.g, sensor input)

1
NEURAL NETWORK REPRESENTATIONS

• A prototypical example of ANN learning is provided by Pomerleau's system ALVINN,

which uses a learned ANN to steer an autonomous vehicle driving at normal speeds on
public highways
• The input to the neural network is a 30x32 grid of pixel intensities obtained from a
forward-pointed camera mounted on the vehicle.
• The network output is the direction in which the vehicle is steered

Figure: Neural network learning to steer an autonomous vehicle.

2
 • Figure illustrates the neural network representation.
• The network is shown on the left side of the figure, with the input camera image depicted
below it.
• Each node (i.e., circle) in the network diagram corresponds to the output of a single
network unit, and the lines entering the node from below are its inputs.
• There are four units that receive inputs directly from all of the 30 x 32 pixels in the
image. These are called "hidden" units because their output is available only within the
network and is not available as part of the global network output. Each of these four
hidden units computes a single real-valued output based on a weighted combination of
its 960 inputs
• These hidden unit outputs are then used as inputs to a second layer of 30 "output" units.
• Each output unit corresponds to a particular steering direction, and the output values of
these units determine which steering direction is recommended most strongly.
• The diagrams on the right side of the figure depict the learned weight values associated
with one of the four hidden units in this ANN.
• The large matrix of black and white boxes on the lower right depicts the weights from
the 30 x 32 pixel inputs into the hidden unit. Here, a white box indicates a positive
weight, a black box a negative weight, and the size of the box indicates the weight
magnitude.
• The smaller rectangular diagram directly above the large matrix shows the weights from
this hidden unit to each of the 30 output units.

APPROPRIATE PROBLEMS FOR NEURAL NETWORK LEARNING

ANN learning is well-suited to problems in which the training data corresponds to noisy,
complex sensor data, such as inputs from cameras and microphones.

ANN is appropriate for problems with the following characteristics:

1. Instances are represented by many attribute-value pairs.

2. The target function output may be discrete-valued, real-valued, or a vector of several
real- or discrete-valued attributes.
3. The training examples may contain errors.
4. Long training times are acceptable.
5. Fast evaluation of the learned target function may be required
6. The ability of humans to understand the learned target function is not important

3
PERCEPTRON

• One type of ANN system is based on a unit called a perceptron. Perceptron is a single
layer neural network.

Figure: A perceptron

• A perceptron takes a vector of real-valued inputs, calculates a linear combination of

these inputs, then outputs a 1 if the result is greater than some threshold and -1 otherwise.
• Given inputs x through x, the output O(x1, . . . , xn) computed by the perceptron is

• Where, each wi is a real-valued constant, or weight, that determines the contribution of

input xi to the perceptron output.
• -w0 is a threshold that the weighted combination of inputs w1x1 + . . . + wnxn must surpass
in order for the perceptron to output a 1.

Sometimes, the perceptron function is written as,

Learning a perceptron involves choosing values for the weights w0 , . . . , wn . Therefore, the
space H of candidate hypotheses considered in perceptron learning is the set of all possible
real-valued weight vectors

4
Representational Power of Perceptrons

• The perceptron can be viewed as representing a hyperplane decision surface in the n-

dimensional space of instances (i.e., points)
• The perceptron outputs a 1 for instances lying on one side of the hyperplane and outputs
a -1 for instances lying on the other side, as illustrated in below figure

Perceptrons can represent all of the primitive Boolean functions AND, OR, NAND (~ AND),
and NOR (~OR)
Some Boolean functions cannot be represented by a single perceptron, such as the XOR
function whose value is 1 if and only if x1 ≠ x2

Example: Representation of AND functions

If A=0 & B=0 → 0*0.6 + 0*0.6 = 0.

This is not greater than the threshold of 1, so the output = 0.
If A=0 & B=1 → 0*0.6 + 1*0.6 = 0.6.
This is not greater than the threshold, so the output = 0.
If A=1 & B=0 → 1*0.6 + 0*0.6 = 0.6.
This is not greater than the threshold, so the output = 0.
If A=1 & B=1 → 1*0.6 + 1*0.6 = 1.2.
This exceeds the threshold, so the output = 1.

5
Drawback of perceptron
• The perceptron rule finds a successful weight vector when the training examples are
linearly separable, it can fail to converge if the examples are not linearly separable

The Perceptron Training Rule

The learning problem is to determine a weight vector that causes the perceptron to produce the
correct + 1 or - 1 output for each of the given training examples.

To learn an acceptable weight vector

• Begin with random weights, then iteratively apply the perceptron to each training
example, modifying the perceptron weights whenever it misclassifies an example.
• This process is repeated, iterating through the training examples as many times as
needed until the perceptron classifies all training examples correctly.
• Weights are modified at each step according to the perceptron training rule, which
revises the weight wi associated with input xi according to the rule.

• The role of the learning rate is to moderate the degree to which weights are changed at
each step. It is usually set to some small value (e.g., 0.1) and is sometimes made to decay
as the number of weight-tuning iterations increases

Drawback:
The perceptron rule finds a successful weight vector when the training examples are linearly
separable, it can fail to converge if the examples are not linearly separable.

6
Gradient Descent and the Delta Rule

• If the training examples are not linearly separable, the delta rule converges toward a
best-fit approximation to the target concept.
• The key idea behind the delta rule is to use gradient descent to search the hypothesis
space of possible weight vectors to find the weights that best fit the training examples.

To understand the delta training rule, consider the task of training an unthresholded perceptron.
That is, a linear unit for which the output O is given by

To derive a weight learning rule for linear units, specify a measure for the training error of a
hypothesis (weight vector), relative to the training examples.

Where,
• D is the set of training examples,
• td is the target output for training example d,
• od is the output of the linear unit for training example d
• E ( w⃗⃗⃗⃗⃗⃗⃗⃗⃗) is simply half the squared difference between the target output td and the linear
unit output od, summed over all training examples.

Visualizing the Hypothesis Space

• To understand the gradient descent algorithm, it is helpful to visualize the entire

hypothesis space of possible weight vectors and their associated E values as shown in
below figure.
• Here the axes w0 and wl represent possible values for the two weights of a simple linear
unit. The w0, wl plane therefore represents the entire hypothesis space.
• The vertical axis indicates the error E relative to some fixed set of training examples.
• The arrow shows the negated gradient at one particular point, indicating the direction in
the w0, wl plane producing steepest descent along the error surface.
• The error surface shown in the figure thus summarizes the desirability of every weight
vector in the hypothesis space

7
 • Given the way in which we chose to define E, for linear units this error surface must
always be parabolic with a single global minimum.

Gradient descent search determines a weight vector that minimizes E by starting with an
arbitrary initial weight vector, then repeatedly modifying it in small steps.
At each step, the weight vector is altered in the direction that produces the steepest descent
along the error surface depicted in above figure. This process continues until the global
minimum error is reached.

Derivation of the Gradient Descent Rule

How to calculate the direction of steepest descent along the error surface?

The direction of steepest can be found by computing the derivative of E with respect to each
component of the vector w⃗⃗⃗⃗⃗⃗⃗⃗⃗. This vector derivative is called the gradient of E with respect to
w⃗⃗⃗⃗⃗⃗⃗⃗⃗, written as

8
The gradient specifies the direction of steepest increase of E, the training rule for
gradient descent is

• Here η is a positive constant called the learning rate, which determines the step
size in the gradient descent search.
• The negative sign is present because we want to move the weight vector in the
direction that decreases E.

This training rule can also be written in its component form

𝜕𝐸
Calculate the gradient at each step. The vector of derivatives that form the
𝜕𝑤𝑖
gradient can be obtained by differentiating E from Equation (2), as

9
GRADIENT DESCENT algorithm for training a linear unit

To summarize, the gradient descent algorithm for training linear units is as follows:
• Pick an initial random weight vector.
• Apply the linear unit to all training examples, then compute Δw i for each weight
according to Equation (7).
• Update each weight wi by adding Δwi, then repeat this process

Issues in Gradient Descent Algorithm

Gradient descent is an important general paradigm for learning. It is a strategy for searching
through a large or infinite hypothesis space that can be applied whenever
1. The hypothesis space contains continuously parameterized hypotheses
2. The error can be differentiated with respect to these hypothesis parameters

The key practical difficulties in applying gradient descent are

1. Converging to a local minimum can sometimes be quite slow
2. If there are multiple local minima in the error surface, then there is no guarantee that
the procedure will find the global minimum

10
Stochastic Approximation to Gradient Descent

• The gradient descent training rule presented in Equation (7) computes weight updates
after summing over all the training examples in D
• The idea behind stochastic gradient descent is to approximate this gradient descent
search by updating weights incrementally, following the calculation of the error for
each individual example
∆wi = η (t – o) xi

• where t, o, and xi are the target value, unit output, and ith input for the training example
in question

One way to view this stochastic gradient descent is to consider a distinct error function
Ed( ⃗⃗⃗w⃗⃗⃗⃗⃗⃗ ) for each individual training example d as follows

• Where, td and od are the target value and the unit output value for training example d.
• Stochastic gradient descent iterates over the training examples d in D, at each iteration
altering the weights according to the gradient with respect to Ed( w⃗⃗⃗⃗⃗⃗⃗⃗⃗)
• The sequence of these weight updates, when iterated over all training examples,
provides a reasonable approximation to descending the gradient with respect to our
original error function Ed( w⃗⃗⃗⃗⃗⃗⃗⃗⃗)
• By making the value of η sufficiently small, stochastic gradient descent can be made to
approximate true gradient descent arbitrarily closely

11
The key differences between standard gradient descent and stochastic gradient descent are
• In standard gradient descent, the error is summed over all examples before updating
weights, whereas in stochastic gradient descent weights are updated upon examining
each training example.
• Summing over multiple examples in standard gradient descent requires more
computation per weight update step. On the other hand, because it uses the true gradient,
standard gradient descent is often used with a larger step size per weight update than
stochastic gradient descent.
• In cases where there are multiple local minima with respect to stochastic gradient
descent can sometimes avoid falling into these local minima because it uses the various
∇E ( ⃗⃗⃗w⃗⃗⃗⃗⃗⃗) rather than ∇ E( w⃗⃗⃗⃗⃗⃗⃗⃗⃗) to guide its search
d

MULTILAYER NETWORKS AND THE BACKPROPAGATION ALGORITHM

Multilayer networks learned by the BACKPROPAGATION algorithm are capable of

expressing a rich variety of nonlinear decision surfaces.

Consider the example:

• Here the speech recognition task involves distinguishing among 10 possible vowels, all
spoken in the context of "h_d" (i.e., "hid," "had," "head," "hood," etc.).
• The network input consists of two parameters, F1 and F2, obtained from a spectral
analysis of the sound. The 10 network outputs correspond to the 10 possible vowel
sounds. The network prediction is the output whose value is highest.
• The plot on the right illustrates the highly nonlinear decision surface represented by the
learned network. Points shown on the plot are test examples distinct from the examples
used to train the network.

12
A Differentiable Threshold Unit (Sigmoid unit)

• Sigmoid unit-a unit very much like a perceptron, but based on a smoothed, differentiable
threshold function.

• The sigmoid unit first computes a linear combination of its inputs, then applies a
threshold to the result and the threshold output is a continuous function of its input.
• More precisely, the sigmoid unit computes its output O as

σ is the sigmoid function

The BACKPROPAGATION Algorithm

• The BACKPROPAGATION Algorithm learns the weights for a multilayer network,

given a network with a fixed set of units and interconnections. It employs gradient
descent to attempt to minimize the squared error between the network output values and
the target values for these outputs.
• In BACKPROPAGATION algorithm, we consider networks with multiple output units
rather than single units as before, so we redefine E to sum the errors over all of the
network output units.

13
where,
• outputs - is the set of output units in the network
• tkd and Okd - the target and output values associated with the kth output unit
• d - training example

Algorithm:

BACK PROPAGATION (training_example, ƞ, nin, nout, nhidden )

Each training example is a pair of the form (𝑥 ⃗⃗⃗ , 𝑡 ), where (𝑥 ) is the vector of network
input values, (𝑡 ) and is the vector of target network output values.
ƞ is the learning rate (e.g., .05). ni, is the number of network inputs, nhidden the number
of units in the hidden layer, and nout the number of output units.
The input from unit i into unit j is denoted xji, and the weight from unit i to unit j is
denoted wji

• Create a feed-forward network with ni inputs, nhidden hidden units, and nout output
units.
• Initialize all network weights to small random numbers
• Until the termination condition is met, Do
• For each (𝑥 ⃗⃗⃗ 𝑡 ), in training examples, Do
,
Propagate the input forward through the network:
1. Input the instance 𝑥⃗⃗⃗, to the network and compute the output ou of
every unit u in the network.
Propagate the errors backward through the network:

14
Adding Momentum
Because BACKPROPAGATION is such a widely used algorithm, many variations have been
developed. The most common is to alter the weight-update rule the equation below

by making the weight update on the nth iteration depend partially on the update that occurred
during the (n - 1)th iteration, as follows:

Learning in arbitrary acyclic networks

• BACKPROPAGATION algorithm given there easily generalizes to feedforward

networks of arbitrary depth. The weight update rule is retained, and the only change is
to the procedure for computing δ values.
• In general, the δ, value for a unit r in layer m is computed from the δ values at the next
deeper layer m + 1 according to

• The rule for calculating δ for any internal unit

Where, Downstream(r) is the set of units immediately downstream from unit r in the network:
that is, all units whose inputs include the output of unit r

Derivation of the BACKPROPAGATION Rule

• Deriving the stochastic gradient descent rule: Stochastic gradient descent involves
iterating through the training examples one at a time, for each training example d
descending the gradient of the error Ed with respect to this single example
• For each training example d every weight wji is updated by adding to it Δwji

15
Here outputs is the set of output units in the network, tk is the target value of unit k for training
example d, and ok is the output of unit k given training example d.

The derivation of the stochastic gradient descent rule is conceptually straightforward, but
requires keeping track of a number of subscripts and variables

• xji = the ith input to unit j

• wji = the weight associated with the ith input to unit j
• netj = Σi wjixji (the weighted sum of inputs for unit j )
• oj = the output computed by unit j
• tj = the target output for unit j
• σ = the sigmoid function
• outputs = the set of units in the final layer of the network
• Downstream(j) = the set of units whose immediate inputs include the output of unit j

16
Consider two cases: The case where unit j is an output unit for the network, and the case where
j is an internal unit (hidden unit).

Case 1: Training Rule for Output Unit Weights.

wji can influence the rest of the network only through netj , netj can influence the network only
through oj. Therefore, we can invoke the chain rule again to write

17
Case 2: Training Rule for Hidden Unit Weights.
• In the case where j is an internal, or hidden unit in the network, the derivation of the
training rule for wji must take into account the indirect ways in which wji can influence
the network outputs and hence Ed.
• For this reason, we will find it useful to refer to the set of all units immediately
downstream of unit j in the network and denoted this set of units by Downstream( j).
• netj can influence the network outputs only through the units in Downstream(j).
Therefore, we can write

18
REMARKS ON THE BACKPROPAGATION ALGORITHM

1. Convergence and Local Minima

• The BACKPROPAGATION multilayer networks is only guaranteed to converge
toward some local minimum in E and not necessarily to the global minimum error.
• Despite the lack of assured convergence to the global minimum error,
BACKPROPAGATION is a highly effective function approximation method in
practice.
• Local minima can be gained by considering the manner in which network weights
evolve as the number of training iterations increases.

Common heuristics to attempt to alleviate the problem of local minima include:

1. Add a momentum term to the weight-update rule. Momentum can sometimes carry the
gradient descent procedure through narrow local minima
2. Use stochastic gradient descent rather than true gradient descent
3. Train multiple networks using the same data, but initializing each network with different
random weights

2. Representational Power of Feedforward Networks

What set of functions can be represented by feed-forward networks?

The answer depends on the width and depth of the networks. There are three quite general
results are known about which function classes can be described by which types of
Networks

1. Boolean functions – Every boolean function can be represented exactly by some

network with two layers of units, although the number of hidden units required grows
exponentially in the worst case with the number of network inputs

2. Continuous functions – Every bounded continuous function can be approximated with

arbitrarily small error by a network with two layers of units

3. Arbitrary functions – Any function can be approximated to arbitrary accuracy by a

network with three layers of units.

3. Hypothesis Space Search and Inductive Bias

• Hypothesis space is the n-dimensional Euclidean space of the n network weights and
hypothesis space is continuous.

19
 • As it is continuous, E is differentiable with respect to the continuous parameters of the
hypothesis, results in a well-defined error gradient that provides a very useful structure
for organizing the search for the best hypothesis.
• It is difficult to characterize precisely the inductive bias of BACKPROPAGATION
algorithm, because it depends on the interplay between the gradient descent search and
the way in which the weight space spans the space of representable functions. However,
one can roughly characterize it as smooth interpolation between data points.

4. Hidden Layer Representations

BACKPROPAGATION can define new hidden layer features that are not explicit in the input
representation, but which capture properties of the input instances that are most relevant to
learning the target function.

Consider example, the network shown in below Figure

20
 • Consider training the network shown in Figure to learn the simple target function f (x)
= x, where x is a vector containing seven 0's and a single 1.
• The network must learn to reproduce the eight inputs at the corresponding eight output
units. Although this is a simple function, the network in this case is constrained to use
only three hidden units. Therefore, the essential information from all eight input units
must be captured by the three learned hidden units.
• When BACKPROPAGATION applied to this task, using each of the eight possible
vectors as training examples, it successfully learns the target function. By examining
the hidden unit values generated by the learned network for each of the eight possible
input vectors, it is easy to see that the learned encoding is similar to the familiar standard
binary encoding of eight values using three bits (e.g., 000,001,010,. . . , 111). The exact
values of the hidden units for one typical run of shown in Figure.
• This ability of multilayer networks to automatically discover useful representations at
the hidden layers is a key feature of ANN learning

5. Generalization, Overfitting, and Stopping Criterion

What is an appropriate condition for terminating the weight update loop? One choice is to
continue training until the error E on the training examples falls below some predetermined
threshold.
To see the dangers of minimizing the error over the training data, consider how the error E
varies with the number of weight iterations

21
• Consider first the top plot in this figure. The lower of the two lines shows the
monotonically decreasing error E over the training set, as the number of gradient descent
iterations grows. The upper line shows the error E measured over a different validation
set of examples, distinct from the training examples. This line measures the
generalization accuracy of the network-the accuracy with which it fits examples beyond
the training data.

• The generalization accuracy measured over the validation examples first decreases, then
increases, even as the error over the training examples continues to decrease. How can
this occur? This occurs because the weights are being tuned to fit idiosyncrasies of the
training examples that are not representative of the general distribution of examples.
The large number of weight parameters in ANNs provides many degrees of freedom for
fitting such idiosyncrasies

• Why does overfitting tend to occur during later iterations, but not during earlier
iterations?
By giving enough weight-tuning iterations, BACKPROPAGATION will often be able
to create overly complex decision surfaces that fit noise in the training data or
unrepresentative characteristics of the particular training sample.

22
 UNIT 3
BAYESIAN AND COMPUTATIONAL LEARNING

Bayesian reasoning provides a probabilistic approach to inference. It is based on the

assumption that the quantities of interest are governed by probability distributions and that
optimal decisions can be made by reasoning about these probabilities together with observed
data

INTRODUCTION

Bayesian learning methods are relevant to study of machine learning for two different reasons.
1. First, Bayesian learning algorithms that calculate explicit probabilities for hypotheses,
such as the naive Bayes classifier, are among the most practical approaches to certain
types of learning problems
2. The second reason is that they provide a useful perspective for understanding many
learning algorithms that do not explicitly manipulate probabilities.

Features of Bayesian Learning Methods

• Each observed training example can incrementally decrease or increase the estimated
probability that a hypothesis is correct. This provides a more flexible approach to
learning than algorithms that completely eliminate a hypothesis if it is found to be
inconsistent with any single example
• Prior knowledge can be combined with observed data to determine the final probability
of a hypothesis. In Bayesian learning, prior knowledge is provided by asserting (1) a
prior probability for each candidate hypothesis, and (2) a probability distribution over
observed data for each possible hypothesis.
• Bayesian methods can accommodate hypotheses that make probabilistic predictions
• New instances can be classified by combining the predictions of multiple hypotheses,
weighted by their probabilities.
• Even in cases where Bayesian methods prove computationally intractable, they can
provide a standard of optimal decision making against which other practical methods
can be measured.

86
Practical difficulty in applying Bayesian methods

1. One practical difficulty in applying Bayesian methods is that they typically require
initial knowledge of many probabilities. When these probabilities are not known in
advance they are often estimated based on background knowledge, previously available
data, and assumptions about the form of the underlying distributions.
2. A second practical difficulty is the significant computational cost required to determine
the Bayes optimal hypothesis in the general case. In certain specialized situations, this
computational cost can be significantly reduced.

BAYES THEOREM

Bayes theorem provides a way to calculate the probability of a hypothesis based on its prior
probability, the probabilities of observing various data given the hypothesis, and the observed
data itself.
Notations
• P(h) prior probability of h, reflects any background knowledge about the chance that h
is correct
• P(D) prior probability of D, probability that D will be observed
• P(D|h) probability of observing D given a world in which h holds
• P(h|D) posterior probability of h, reflects confidence that h holds after D has been
observed

Bayes theorem is the cornerstone of Bayesian learning methods because it provides a way to
calculate the posterior probability P(h|D), from the prior probability P(h), together with P(D)
and P(D|h).

• P(h|D) increases with P(h) and with P(D|h) according to Bayes theorem.
• P(h|D) decreases as P(D) increases, because the more probable it is that D will be
observed independent of h, the less evidence D provides in support of h.

87
Maximum a Posteriori (MAP) Hypothesis

• In many learning scenarios, the learner considers some set of candidate hypotheses H
and is interested in finding the most probable hypothesis h ∈ H given the observed data
D. Any such maximally probable hypothesis is called a maximum a posteriori (MAP)
hypothesis.
• Bayes theorem to calculate the posterior probability of each candidate hypothesis is hMAP
is a MAP hypothesis provided

• P(D) can be dropped, because it is a constant independent of h

Maximum Likelihood (ML) Hypothesis

• In some cases, it is assumed that every hypothesis in H is equally probable a priori

(P(hi) = P(hj) for all hi and hj in H).
• In this case the below equation can be simplified and need only consider the term P(D|h)
to find the most probable hypothesis.

P(D|h) is often called the likelihood of the data D given h, and any hypothesis that maximizes
P(D|h) is called a maximum likelihood (ML) hypothesis

Example
• Consider a medical diagnosis problem in which there are two alternative hypotheses:
(1) that the patient has particular form of cancer, and (2) that the patient does not. The
available data is from a particular laboratory test with two possible outcomes: +
(positive) and - (negative).

88
 • We have prior knowledge that over the entire population of people only .008 have this
disease. Furthermore, the lab test is only an imperfect indicator of the disease.
• The test returns a correct positive result in only 98% of the cases in which the disease is
actually present and a correct negative result in only 97% of the cases in which the
disease is not present. In other cases, the test returns the opposite result.
• The above situation can be summarized by the following probabilities:

Suppose a new patient is observed for whom the lab test returns a positive (+) result.
Should we diagnose the patient as having cancer or not?

The exact posterior probabilities can also be determined by normalizing the above quantities
so that they sum to 1

Basic formulas for calculating probabilities are summarized in Table

89
BAYES THEOREM AND CONCEPT LEARNING

What is the relationship between Bayes theorem and the problem of concept learning?

Since Bayes theorem provides a principled way to calculate the posterior probability of each
hypothesis given the training data, and can use it as the basis for a straightforward learning
algorithm that calculates the probability for each possible hypothesis, then outputs the most
probable.

Brute-Force Bayes Concept Learning

Consider the concept learning problem

• Assume the learner considers some finite hypothesis space H defined over the instance
space X, in which the task is to learn some target concept c : X → {0,1}.
• Learner is given some sequence of training examples ((x1, d1) . . . (xm, dm)) where xi is
some instance from X and where di is the target value of xi (i.e., di = c(xi)).
• The sequence of target values are written as D = (d1 . . . dm).

We can design a straightforward concept learning algorithm to output the maximum a posteriori
hypothesis, based on Bayes theorem, as follows:

BRUTE-FORCE MAP LEARNING algorithm:

1. For each hypothesis h in H, calculate the posterior probability

2. Output the hypothesis hMAP with the highest posterior probability

In order specify a learning problem for the BRUTE-FORCE MAP LEARNING algorithm we
must specify what values are to be used for P(h) and for P(D|h) ?

Let’s choose P(h) and for P(D|h) to be consistent with the following assumptions:
• The training data D is noise free (i.e., di = c(xi))
• The target concept c is contained in the hypothesis space H
• Do not have a priori reason to believe that any hypothesis is more probable than any
other.

90
What values should we specify for P(h)?
• Given no prior knowledge that one hypothesis is more likely than another, it is
reasonable to assign the same prior probability to every hypothesis h in H.
• Assume the target concept is contained in H and require that these prior probabilities
sum to 1.

What choice shall we make for P(D|h)?

• P(D|h) is the probability of observing the target values D = (d1 . . .dm) for the fixed set
of instances (x1 . . . xm), given a world in which hypothesis h holds
• Since we assume noise-free training data, the probability of observing classification di
given h is just 1 if di = h(xi) and 0 if di ≠ h(xi). Therefore,

Given these choices for P(h) and for P(D|h) we now have a fully-defined problem for the above
BRUTE-FORCE MAP LEARNING algorithm.

Recalling Bayes theorem, we have

Consider the case where h is inconsistent with the training data D

The posterior probability of a hypothesis inconsistent with D is zero

Consider the case where h is consistent with D

Where, VSH,D is the subset of hypotheses from H that are consistent with D

To summarize, Bayes theorem implies that the posterior probability P(h|D) under our assumed
P(h) and P(D|h) is

91
The Evolution of Probabilities Associated with Hypotheses

• Figure (a) all hypotheses have the same probability.

• Figures (b) and (c), As training data accumulates, the posterior probability for
inconsistent hypotheses becomes zero while the total probability summing to 1 is
shared equally among the remaining consistent hypotheses.

MAP Hypotheses and Consistent Learners

• A learning algorithm is a consistent learner if it outputs a hypothesis that commits zero

errors over the training examples.
• Every consistent learner outputs a MAP hypothesis, if we assume a uniform prior
probability distribution over H (P(hi) = P(hj) for all i, j), and deterministic, noise free
training data (P(D|h) =1 if D and h are consistent, and 0 otherwise).

Example:
• FIND-S outputs a consistent hypothesis, it will output a MAP hypothesis under the
probability distributions P(h) and P(D|h) defined above.
• Are there other probability distributions for P(h) and P(D|h) under which FIND-S
outputs MAP hypotheses? Yes.
• Because FIND-S outputs a maximally specific hypothesis from the version space, its
output hypothesis will be a MAP hypothesis relative to any prior probability distribution
that favours more specific hypotheses.

Note
• Bayesian framework is a way to characterize the behaviour of learning algorithms
• By identifying probability distributions P(h) and P(D|h) under which the output is a
optimal hypothesis, implicit assumptions of the algorithm can be characterized
(Inductive Bias)
• Inductive inference is modelled by an equivalent probabilistic reasoning system based
on Bayes theorem

92
MAXIMUM LIKELIHOOD AND LEAST-SQUARED ERROR HYPOTHESES

Consider the problem of learning a continuous-valued target function such as neural network
learning, linear regression, and polynomial curve fitting

A straightforward Bayesian analysis will show that under certain assumptions any learning
algorithm that minimizes the squared error between the output hypothesis predictions and the
training data will output a maximum likelihood (ML) hypothesis

• Learner L considers an instance space X and a hypothesis space H consisting of some

class of real-valued functions defined over X, i.e., (∀ h ∈ H)[ h : X → R] and training
examples of the form <xi,di>
• The problem faced by L is to learn an unknown target function f : X → R
• A set of m training examples is provided, where the target value of each example is
corrupted by random noise drawn according to a Normal probability distribution with
zero mean (di = f(xi) + ei)
• Each training example is a pair of the form (xi ,di ) where di = f (xi ) + ei .
– Here f(xi) is the noise-free value of the target function and ei is a random variable
representing the noise.
– It is assumed that the values of the ei are drawn independently and that they are
distributed according to a Normal distribution with zero mean.
• The task of the learner is to output a maximum likelihood hypothesis or a MAP
hypothesis assuming all hypotheses are equally probable a priori.

Using the definition of hML we have

Assuming training examples are mutually independent given h, we can write P(D|h) as the
product of the various (di|h)

Given the noise ei obeys a Normal distribution with zero mean and unknown variance σ2 , each
di must also obey a Normal distribution around the true targetvalue f(xi). Because we are
writing the expression for P(D|h), we assume h is the correct description of f.
Hence, µ = f(xi) = h(xi)

93
Maximize the less complicated logarithm, which is justified because of the monotonicity of
function p

The first term in this expression is a constant independent of h, and can therefore be
discarded, yielding

Maximizing this negative quantity is equivalent to minimizing the corresponding positive

quantity

Finally, discard constants that are independent of h.

Thus, above equation shows that the maximum likelihood hypothesis hML is the one that
minimizes the sum of the squared errors between the observed training values d i and the
hypothesis predictions h(xi)

Note:
Why is it reasonable to choose the Normal distribution to characterize noise?
• Good approximation of many types of noise in physical systems
• Central Limit Theorem shows that the sum of a sufficiently large number of
independent, identically distributed random variables itself obeys a Normal distribution
Only noise in the target value is considered, not in the attributes describing the instances
themselves

94
MAXIMUM LIKELIHOOD HYPOTHESES FOR PREDICTING PROBABILITIES

• Consider the setting in which we wish to learn a nondeterministic (probabilistic)

function f : X → {0, 1}, which has two discrete output values.
• We want a function approximator whose output is the probability that f(x) = 1. In other
words, learn the target function f ` : X → [0, 1] such that f ` (x) = P(f(x) = 1)

How can we learn f ` using a neural network?

• Use of brute force way would be to first collect the observed frequencies of 1's and 0's
for each possible value of x and to then train the neural network to output the target
frequency for each x.

What criterion should we optimize in order to find a maximum likelihood hypothesis for f' in
this setting?
• First obtain an expression for P(D|h)
• Assume the training data D is of the form D = {(x1, d1) . . . (xm, dm)}, where di is the
observed 0 or 1 value for f (xi).
• Both xi and di as random variables, and assuming that each training example is drawn
independently, we can write P(D|h) as

Applying the product rule

The probability P(di|h, xi)

Re-express it in a more mathematically manipulable form, as

Equation (4) to substitute for P(di |h, xi) in Equation (5) to obtain

We write an expression for the maximum likelihood hypothesis

95
The last term is a constant independent of h, so it can be dropped

It easier to work with the log of the likelihood, yielding

Equation (7) describes the quantity that must be maximized in order to obtain the maximum
likelihood hypothesis in our current problem setting

Gradient Search to Maximize Likelihood in a Neural Net

• Derive a weight-training rule for neural network learning that seeks to maximize G(h,D)
using gradient ascent
• The gradient of G(h,D) is given by the vector of partial derivatives of G(h,D) with
respect to the various network weights that define the hypothesis h represented by the
learned network
• In this case, the partial derivative of G(h, D) with respect to weight wjk from input k to
unit j is

• Suppose our neural network is constructed from a single layer of sigmoid units. Then,

where xijk is the kth input to unit j for the ith training example, and d(x) is the derivative
of the sigmoid squashing function.

• Finally, substituting this expression into Equation (1), we obtain a simple expression for
the derivatives that constitute the gradient

96
Because we seek to maximize rather than minimize P(D|h), we perform gradient ascent rather
than gradient descent search. On each iteration of the search the weight vector is adjusted in
the direction of the gradient, using the weight update rule

Where, η is a small positive constant that determines the step size of the i gradient ascent search

MINIMUM DESCRIPTION LENGTH PRINCIPLE

• A Bayesian perspective on Occam’s razor

• Motivated by interpreting the definition of hMAP in the light of basic concepts from
information theory.

which can be equivalently expressed in terms of maximizing the log2

or alternatively, minimizing the negative of this quantity

This equation (1) can be interpreted as a statement that short hypotheses are preferred,
assuming a particular representation scheme for encoding hypotheses and data

• -log2P(h): the description length of h under the optimal encoding for the hypothesis
space H, LCH (h) = −log2P(h), where CH is the optimal code for hypothesis space H.
• -log2P(D | h): the description length of the training data D given hypothesis h, under the
optimal encoding from the hypothesis space H: LCH (D|h) = −log2P(D| h) , where C D|h
is the optimal code for describing data D assuming that both the sender and receiver
know the hypothesis h.
• Rewrite Equation (1) to show that hMAP is the hypothesis h that minimizes the sum given
by the description length of the hypothesis plus the description length of the data given
the hypothesis.

Where, CH and CD|h are the optimal encodings for H and for D given h

97
The Minimum Description Length (MDL) principle recommends choosing the hypothesis that
minimizes the sum of these two description lengths of equ.

Minimum Description Length principle:

Where, codes C1 and C2 to represent the hypothesis and the data given the hypothesis

The above analysis shows that if we choose C1 to be the optimal encoding of hypotheses CH,
and if we choose C2 to be the optimal encoding CD|h, then hMDL = hMAP

Application to Decision Tree Learning

Apply the MDL principle to the problem of learning decision trees from some training data.
What should we choose for the representations C1 and C2 of hypotheses and data?
• For C1: C1 might be some obvious encoding, in which the description length grows with
the number of nodes and with the number of edges
• For C2: Suppose that the sequence of instances (x1 . . .xm) is already known to both the
transmitter and receiver, so that we need only transmit the classifications (f (x1) . . . f
(xm)).
• Now if the training classifications (f (x1) . . .f(xm)) are identical to the predictions of the
hypothesis, then there is no need to transmit any information about these examples. The
description length of the classifications given the hypothesis ZERO
• If examples are misclassified by h, then for each misclassification we need to transmit
a message that identifies which example is misclassified as well as its correct
classification
• The hypothesis hMDL under the encoding C1 and C2 is just the one that minimizes the
sum of these description lengths.

98
NAIVE BAYES CLASSIFIER

• The naive Bayes classifier applies to learning tasks where each instance x is described
by a conjunction of attribute values and where the target function f (x) can take on any
value from some finite set V.
• A set of training examples of the target function is provided, and a new instance is
presented, described by the tuple of attribute values (al, a2.. .am).
• The learner is asked to predict the target value, or classification, for this new instance.

The Bayesian approach to classifying the new instance is to assign the most probable target
value, VMAP, given the attribute values (al, a2.. .am) that describe the instance

Use Bayes theorem to rewrite this expression as

• The naive Bayes classifier is based on the assumption that the attribute values are
conditionally independent given the target value. Means, the assumption is that given
the target value of the instance, the probability of observing the conjunction (al, a2.. .am),
is just the product of the probabilities for the individual attributes:

Substituting this into Equation (1),

Naive Bayes classifier:

Where, VNB denotes the target value output by the naive Bayes classifier

99
An Illustrative Example
• Let us apply the naive Bayes classifier to a concept learning problem i.e., classifying
days according to whether someone will play tennis.
• The below table provides a set of 14 training examples of the target concept PlayTennis,
where each day is described by the attributes Outlook, Temperature, Humidity, and
Wind

Day Outlook Temperature Humidity Wind PlayTennis

D1 Sunny Hot High Weak No
D2 Sunny Hot High Strong No
D3 Overcast Hot High Weak Yes
D4 Rain Mild High Weak Yes
D5 Rain Cool Normal Weak Yes
D6 Rain Cool Normal Strong No
D7 Overcast Cool Normal Strong Yes
D8 Sunny Mild High Weak No
D9 Sunny Cool Normal Weak Yes
D10 Rain Mild Normal Weak Yes
D11 Sunny Mild Normal Strong Yes
D12 Overcast Mild High Strong Yes
D13 Overcast Hot Normal Weak Yes
D14 Rain Mild High Strong No

• Use the naive Bayes classifier and the training data from this table to classify the
following novel instance:
< Outlook = sunny, Temperature = cool, Humidity = high, Wind = strong >

• Our task is to predict the target value (yes or no) of the target concept PlayTennis for
this new instance

100
The probabilities of the different target values can easily be estimated based on their
frequencies over the 14 training examples
• P(P1ayTennis = yes) = 9/14 = 0.64
• P(P1ayTennis = no) = 5/14 = 0.36

Similarly, estimate the conditional probabilities. For example, those for Wind = strong
• P(Wind = strong | PlayTennis = yes) = 3/9 = 0.33
• P(Wind = strong | PlayTennis = no) = 3/5 = 0.60

Calculate VNB according to Equation (1)

Thus, the naive Bayes classifier assigns the target value PlayTennis = no to this new
instance, based on the probability estimates learned from the training data.

By normalizing the above quantities to sum to one, calculate the conditional probability that
the target value is no, given the observed attribute values

Estimating Probabilities

• We have estimated probabilities by the fraction of times the event is observed to occur
over the total number of opportunities.
• For example, in the above case we estimated P(Wind = strong | Play Tennis = no) by
the fraction nc /n where, n = 5 is the total number of training examples for which
PlayTennis = no, and nc = 3 is the number of these for which Wind = strong.
• When nc = 0, then nc /n will be zero and this probability term will dominate the quantity
calculated in Equation (2) requires multiplying all the other probability terms by this
zero value
• To avoid this difficulty we can adopt a Bayesian approach to estimating the probability,
using the m-estimate defined as follows

m -estimate of probability:

101
 • p is our prior estimate of the probability we wish to determine, and m is a constant
called the equivalent sample size, which determines how heavily to weight p relative
to the observed data
• Method for choosing p in the absence of other information is to assume uniform
priors; that is, if an attribute has k possible values we set p = 1 /k.

BAYESIAN BELIEF NETWORKS

• The naive Bayes classifier makes significant use of the assumption that the values of the
attributes a1 . . .an are conditionally independent given the target value v.
• This assumption dramatically reduces the complexity of learning the target function

A Bayesian belief network describes the probability distribution governing a set of variables
by specifying a set of conditional independence assumptions along with a set of conditional
probabilities
Bayesian belief networks allow stating conditional independence assumptions that apply to
subsets of the variables

Notation
• Consider an arbitrary set of random variables Y1 . . . Yn , where each variable Yi can
take on the set of possible values V(Yi).
• The joint space of the set of variables Y to be the cross product V(Y1) x V(Y2) x. . .
V(Yn).
• In other words, each item in the joint space corresponds to one of the possible
assignments of values to the tuple of variables (Y1 . . . Yn). The probability distribution
over this joint' space is called the joint probability distribution.
• The joint probability distribution specifies the probability for each of the possible
variable bindings for the tuple (Y1 . . . Yn).
• A Bayesian belief network describes the joint probability distribution for a set of
variables.

Conditional Independence

Let X, Y, and Z be three discrete-valued random variables. X is conditionally independent of

Y given Z if the probability distribution governing X is independent of the value of Y given a
value for Z, that is, if

Where,

102
The above expression is written in abbreviated form as
P(X | Y, Z) = P(X | Z)

Conditional independence can be extended to sets of variables. The set of variables X 1 . . . Xl

is conditionally independent of the set of variables Y1 . . . Ym given the set of variables Z1 . . .
Zn if

The naive Bayes classifier assumes that the instance attribute A1 is conditionally independent
of instance attribute A2 given the target value V. This allows the naive Bayes classifier to
calculate P(Al, A2 | V) as follows,

Representation

A Bayesian belief network represents the joint probability distribution for a set of variables.
Bayesian networks (BN) are represented by directed acyclic graphs.

The Bayesian network in above figure represents the joint probability distribution over the
boolean variables Storm, Lightning, Thunder, ForestFire, Campfire, and BusTourGroup

A Bayesian network (BN) represents the joint probability distribution by specifying a set of
conditional independence assumptions
• BN represented by a directed acyclic graph, together with sets of local conditional
probabilities
• Each variable in the joint space is represented by a node in the Bayesian network
• The network arcs represent the assertion that the variable is conditionally independent
of its non-descendants in the network given its immediate predecessors in the network.
• A conditional probability table (CPT) is given for each variable, describing the
probability distribution for that variable given the values of its immediate predecessors

103
The joint probability for any desired assignment of values (y1, . . . , yn) to the tuple of network
variables (Y1 . . . Ym) can be computed by the formula

Where, Parents(Yi) denotes the set of immediate predecessors of Yi in the network.

Example:
Consider the node Campfire. The network nodes and arcs represent the assertion that Campfire
is conditionally independent of its non-descendants Lightning and Thunder, given its
immediate parents Storm and BusTourGroup.

This means that once we know the value of the variables Storm and BusTourGroup, the
variables Lightning and Thunder provide no additional information about Campfire
The conditional probability table associated with the variable Campfire. The assertion is

P(Campfire = True | Storm = True, BusTourGroup = True) = 0.4

Inference

• Use a Bayesian network to infer the value of some target variable (e.g., ForestFire) given
the observed values of the other variables.
• Inference can be straightforward if values for all of the other variables in the network
are known exactly.
• A Bayesian network can be used to compute the probability distribution for any subset
of network variables given the values or distributions for any subset of the remaining
variables.
• An arbitrary Bayesian network is known to be NP-hard

104
Learning Bayesian Belief Networks

Affective algorithms can be considered for learning Bayesian belief networks from training
data by considering several different settings for learning problem
➢ First, the network structure might be given in advance, or it might have to be inferred from
the training data.
➢ Second, all the network variables might be directly observable in each training example,
or some might be unobservable.
• In the case where the network structure is given in advance and the variables are fully
observable in the training examples, learning the conditional probability tables is
straightforward and estimate the conditional probability table entries
• In the case where the network structure is given but only some of the variable values
are observable in the training data, the learning problem is more difficult. The learning
problem can be compared to learning weights for an ANN.

Gradient Ascent Training of Bayesian Network

The gradient ascent rule which maximizes P(D|h) by following the gradient of ln P(D|h) with
respect to the parameters that define the conditional probability tables of the Bayesian network.

Let wijk denote a single entry in one of the conditional probability tables. In particular wijk
denote the conditional probability that the network variable Yi will take on the value yi, given
that its immediate parents Ui take on the values given by uik.

The gradient of ln P(D|h) is given by the derivatives for each of the wijk.
As shown below, each of these derivatives can be calculated as

Derive the gradient defined by the set of derivatives for all i, j, and k. Assuming the
training examples d in the data set D are drawn independently, we write this derivative as

105
We write the abbreviation Ph(D) to represent P(D|h).

106
THE EM ALGORITHM

The EM algorithm can be used even for variables whose value is never directly observed,
provided the general form of the probability distribution governing these variables is known.

Estimating Means of k Gaussians

• Consider a problem in which the data D is a set of instances generated by a probability

distribution that is a mixture of k distinct Normal distributions.

• This problem setting is illustrated in Figure for the case where k = 2 and where the
instances are the points shown along the x axis.
• Each instance is generated using a two-step process.
• First, one of the k Normal distributions is selected at random.
• Second, a single random instance xi is generated according to this selected
distribution.
• This process is repeated to generate a set of data points as shown in the figure.

107
 • To simplify, consider the special case
• The selection of the single Normal distribution at each step is based on choosing
each with uniform probability
• Each of the k Normal distributions has the same variance σ2, known value.
• The learning task is to output a hypothesis h = (μ1 , . . . ,μk) that describes the means of
each of the k distributions.
• We would like to find a maximum likelihood hypothesis for these means; that is, a
hypothesis h that maximizes p(D |h).

In this case, the sum of squared errors is minimized by the sample mean

• Our problem here, however, involves a mixture of k different Normal distributions, and
we cannot observe which instances were generated by which distribution.
• Consider full description of each instance as the triple (xi, zi1, zi2),
• where xi is the observed value of the ith instance and
• where zi1 and zi2 indicate which of the two Normal distributions was used to
generate the value xi
• In particular, zij has the value 1 if xi was created by the jth Normal distribution and 0
otherwise.
• Here xi is the observed variable in the description of the instance, and zil and zi2 are
hidden variables.
• If the values of zil and zi2 were observed, we could use following Equation to solve for
the means p1 and p2
• Because they are not, we will instead use the EM algorithm

EM algorithm

108
109