IAT-II Question Paper With Solution of 18CS71 Artificial Intelligence and Machine Learning Dec-2021-Dr - Kavitha P and MR - Radha Krishnan
IAT-II Question Paper With Solution of 18CS71 Artificial Intelligence and Machine Learning Dec-2021-Dr - Kavitha P and MR - Radha Krishnan
IAT-II Question Paper With Solution of 18CS71 Artificial Intelligence and Machine Learning Dec-2021-Dr - Kavitha P and MR - Radha Krishnan
Answer:
Entropy of Dataset = - [(9/4) log (9/14) + (5/14). Log (5/14) = 0.941
Entropy(S,outlook) = - (5/14) * [ (2/5) log (2/5) + (3/5) log(3/5)
- 4/14 * 0
- (5/14) * [ (2/5) log (2/5) + (3/5) log(3/5)
= 0.693
Information Gain(S,outlook) = 0.940- 0.693 = 0.246
Entropy(S, Temp) = - (4/14) * -1
- (6/14) *[ (2/6) log(2/6) + (4/6) log (4/6)]
- (4/14) * [ (1/4)log(1/4) + (3/4) log(3/4)]
= 0.911
Information Gain(S, Temp = 0.940 = 0.911 = 0.029
Entropy(S, Humidity) = - (7/14) * [ (4/7) log(4/7) + (3/7) log(3/7)]
- (7/14) * [ (6/7) log(6/7) + (1/7) log(1/7)
= 0.789
Information Gain(S, Humidity) = 0.940-0.789 = 0.151
Entropy(S, Wind) = - (6/14) * (-1)
- (8/14) * [ (6/8) log(6/8) + (2/8) log(2/8)]
= 0.892
Info Gain(S, Wind) = 0.940-0.892 = 0.048
Attribute “Outlook” has maximum info gain. Hence root of the tree will be
“Outlook”
Nodes at next Level
1. “Sunny” branch
Humidity divides the subset of samples perfectly and has entropy of 0. Hence
humidity will be the node at the “Sunny” branch
2. “Overcast” branch
All samples are “Yes” class. Hence a leaf node with “Yes” is at this level
3. “Rain” branch
Wind divides the subset of samples perfectly and has entropy of 0. Hence Wind
will be the node at the “Rain” branch
Nodes at level 3
1. (Outlook=Sunny, Humidity=High) branch
Leaf node “No”
2. (Outlook=Sunny, Humidity=Normal) branch
Leaf node “Yes”
3. (Outlook=Rain, Wind=Weak) branch
Leaf node “Yes”
4. (Outlook=Rain, Wind=Strong) branch
Leaf node “No”
2 (2) Construct the decision tree to represent the following Boolean [05] CO2 L2
ii) A∨ [B∧ C]
2 (b) How a single perceptron can be used to represent the Boolean functions [05] CO2 L2
such as AND, OR.
2.5+2.5
3 Explain the back propagation algorithm. Why is it not likely to be [7+3] CO2 L2
trapped in local minima?
7+3
Large number of dimensions can provide an escape route from local
minima. The sigmoid function which has small slope for large positive
and negative values of linear sum also helps in avoiding local minima in
most cases.
4 (a) Write Gradient descent algorithm to train a linear unit along with the [05] CO2 L2
derivation
2.5+2.5
Derivation
4 (b) Write the characteristics of the problems appropriate for solving using [05] CO1 L1
neural network.
ANN is appropriate for problems with the following characteristics 05 marks –
one per
•Instances are represented by many attribute value pairs. characteristi
The ta r g e t f unction o u t p ut m ay be discr e t e v a lued, cs
rea l v alued, or a ve c tor of several real or discrete valued
attributes.
•The training examples may contain errors.
•Long training times are acceptable.
•Fast evaluation of the learned target function may be required
•The ability of humans to understand the learned target function
is not important
5 Show the derivation of back propagation training rule for output unit [10] CO2 L2
weights.
Δ wji
6(a) Explain the various approaches for knowledge representation [05] CO1 L1
Modules
covered
Blooms
PSO1
PSO2
PSO3
PSO4
PO10
PO11
PO12
Level
PO1
PO2
PO3
PO4
PO5
PO6
PO7
PO8
PO9
Course Outcomes
COGNITIVE
REVISED BLOOMS TAXONOMY KEYWORDS
LEVEL
List, define, tell, describe, identify, show, label, collect, examine, tabulate, quote, name, who,
L1
when, where, etc.
summarize, describe, interpret, contrast, predict, associate, distinguish, estimate, differentiate,
L2
discuss, extend
Apply, demonstrate, calculate, complete, illustrate, show, solve, examine, modify, relate,
L3
change, classify, experiment, discover.
Analyze, separate, order, explain, connect, classify, arrange, divide, compare, select, explain,
L4
infer.
Assess, decide, rank, grade, test, measure, recommend, convince, select, judge, explain,
L5
discriminate, support, conclude, compare, summarize.
CORRELATION
PROGRAM OUTCOMES (PO), PROGRAM SPECIFIC OUTCOMES (PSO)
LEVELS
PO1 Engineering knowledge PO7 Environment and sustainability 0 No Correlation
PO2 Problem analysis PO8 Ethics 1 Slight/Low
Moderate/
PO3 Design/development of solutions PO9 Individual and team work 2
Medium
Conduct investigations of Substantial/
PO4 PO10 Communication 3
complex problems High
PO5 Modern tool usage PO11 Project management and finance
PO6 The Engineer and society PO12 Life-long learning
PSO1 Develop applications using different stacks of web and programming technologies
PSO2 Design and develop secure, parallel, distributed, networked, and digital systems
PSO3 Apply software engineering methods to design, develop, test and manage software systems.
PSO4 Develop intelligent applications for business and industry