07/01/2021 Machine Learning - Home | Coursera

Week 2 Lecture Notes

ML:Linear Regression with Multiple Variables

Linear regression with multiple variables is also known as "multivariate linear regression".

We now introduce notation for equations where we can have any number of input variables.

Now define the multivariable form of the hypothesis function as follows, accommodating these multiple features:

hθ (x) = θ0 + θ1 x1 + θ2 x2 + θ3 x3 + ⋯ + θn xn

In order to develop intuition about this function, we can think about θ0 as the basic price of a house, θ1 as the price per square meter, θ2 as the
price per floor, etc. x1 will be the number of square meters in the house, x2 the number of floors, etc.

Using the definition of matrix multiplication, our multivariable hypothesis function can be concisely represented as:

This is a vectorization of our hypothesis function for one training example; see the lessons on vectorization to learn more.

(i)
Remark: Note that for convenience reasons in this course Mr. Ng assumes x0 = 1 for (i ∈ 1, … , m)

(i)
[Note: So that we can do matrix operations with theta and x, we will set x0 = 1, for all values of i. This makes the two vectors 'theta' and x(i)
match each other element-wise (that is, have the same number of elements: n+1).]

The training examples are stored in X row-wise, like such:

You can calculate the hypothesis as a column vector of size (m x 1) with:

hθ (X) = Xθ

For the rest of these notes, and other lecture notes, X will represent a matrix of training examples x(i) stored row-wise.

Cost function
For the parameter vector θ (of type Rn+1 or in R(n+1)×1 , the cost function is:

m
1 2
J(θ) = ∑ (hθ (x(i) ) − y (i) )
2m i=1

The vectorized version is:

1
J(θ) = (Xθ − y )T (Xθ − y )
2m

Where y denotes the vector of all y values.

Gradient Descent for Multiple Variables

https://www.coursera.org/learn/machine-learning/resources/QQx8l 1/680
07/01/2021 Machine Learning - Home | Coursera

The gradient descent equation itself is generally the same form; we just have to repeat it for our 'n' features:

In other words:

Matrix Notation
The Gradient Descent rule can be expressed as:

θ := θ − α∇J(θ)

Where ∇J(θ) is a column vector of the form:

∇J(θ) = [ ∂J(θ) ∂J(θ) ⋮ ∂J(θ) ]

∂θ0 ∂θ1 ∂θn

The j-th component of the gradient is the summation of the product of two terms:

Sometimes, the summation of the product of two terms can be expressed as the product of two vectors.

(i)
Here, xj , for i = 1,...,m, represents the m elements of the j-th column, xj , of the training set X.

∂J(θ)
The other term (hθ (x(i) ) − y (i) ) is the vector of the deviations between the predictions hθ (x(i) ) and the true values y (i) . Re-writing ∂θ , we
j
have:

Finally, the matrix notation (vectorized) of the Gradient Descent rule is:

θ := θ − α
m
X T (Xθ − y)

Feature Normalization
We can speed up gradient descent by having each of our input values in roughly the same range. This is because θ will descend quickly on small
ranges and slowly on large ranges, and so will oscillate inefficiently down to the optimum when the variables are very uneven.

The way to prevent this is to modify the ranges of our input variables so that they are all roughly the same. Ideally:

−1 ≤ x(i) ≤ 1

or

−0.5 ≤ x(i) ≤ 0.5

https://www.coursera.org/learn/machine-learning/resources/QQx8l 2/680
07/01/2021 Machine Learning - Home | Coursera

These aren't exact requirements; we are only trying to speed things up. The goal is to get all input variables into roughly one of these ranges, give
or take a few.

Two techniques to help with this are feature scaling and mean normalization. Feature scaling involves dividing the input values by the range (i.e.
the maximum value minus the minimum value) of the input variable, resulting in a new range of just 1. Mean normalization involves subtracting
the average value for an input variable from the values for that input variable, resulting in a new average value for the input variable of just zero. To
implement both of these techniques, adjust your input values as shown in this formula:

xi − μi
xi :=
si

Where μi is the average of all the values for feature (i) and si is the range of values (max - min), or si is the standard deviation.

Note that dividing by the range, or dividing by the standard deviation, give different results. The quizzes in this course use range - the programming
exercises use standard deviation.

price − 1000
Example: xi is housing prices with range of 100 to 2000, with a mean value of 1000. Then, xi := .
1900

Quiz question #1 on Feature Normalization (Week 2, Linear Regression with Multiple

Variables)
Your answer should be rounded to exactly two decimal places. Use a '.' for the decimal point, not a ','. The tricky part of this question is figuring out
which feature of which training example you are asked to normalize. Note that the mobile app doesn't allow entering a negative number (Jan 2016),
so you will need to use a browser to submit this quiz if your solution requires a negative number.

Gradient Descent Tips

Debugging gradient descent. Make a plot with number of iterations on the x-axis. Now plot the cost function, J(θ) over the number of iterations of
gradient descent. If J(θ) ever increases, then you probably need to decrease α.

Automatic convergence test. Declare convergence if J(θ) decreases by less than E in one iteration, where E is some small value such as 10−3.
However in practice it's difficult to choose this threshold value.

It has been proven that if learning rate α is sufficiently small, then J(θ) will decrease on every iteration. Andrew Ng recommends decreasing α by
multiples of 3.

Features and Polynomial Regression

We can improve our features and the form of our hypothesis function in a couple different ways.

We can combine multiple features into one. For example, we can combine x1 and x2 into a new feature x3 by taking x1 ⋅x2 .

Polynomial Regression

Our hypothesis function need not be linear (a straight line) if that does not fit the data well.

We can change the behavior or curve of our hypothesis function by making it a quadratic, cubic or square root function (or any other form).

For example, if our hypothesis function is hθ (x) = θ0 + θ1 x1 then we can create additional features based on x1 , to get the quadratic function
hθ (x) = θ0 + θ1 x1 + θ2 x21 or the cubic function hθ (x) = θ0 + θ1 x1 + θ2 x21 + θ3 x31

In the cubic version, we have created new features x2 and x3 where x2 = x21 and x3 = x31 .

To make it a square root function, we could do: hθ (x) = θ0 + θ1 x1 + θ2 x1

Note that at 2:52 and through 6:22 in the "Features and Polynomial Regression" video, the curve that Prof Ng discusses about "doesn't ever come
back down" is in reference to the hypothesis function that uses the sqrt() function (shown by the solid purple line), not the one that uses size2
(shown with the dotted blue line). The quadratic form of the hypothesis function would have the shape shown with the blue dotted line if θ2 was
negative.

One important thing to keep in mind is, if you choose your features this way then feature scaling becomes very important.

eg. if x1 has range 1 - 1000 then range of x21 becomes 1 - 1000000 and that of x31 becomes 1 - 1000000000.

Normal Equation
The "Normal Equation" is a method of finding the optimum theta without iteration.

https://www.coursera.org/learn/machine-learning/resources/QQx8l 3/680
07/01/2021 Machine Learning - Home | Coursera

θ = (X T X)−1 X T y

There is no need to do feature scaling with the normal equation.

Mathematical proof of the Normal equation requires knowledge of linear algebra and is fairly involved, so you do not need to worry about the
details.

Proofs are available at these links for those who are interested:

https://en.wikipedia.org/wiki/Linear_least_squares_(mathematics)

http://eli.thegreenplace.net/2014/derivation-of-the-normal-equation-for-linear-regression

The following is a comparison of gradient descent and the normal equation:

Gradient Descent Normal Equation

Need to choose alpha No need to choose alpha

Needs many iterations No need to iterate

2
O (kn ) O (n3 ), need to calculate inverse of X T X

Works well when n is large Slow if n is very large

With the normal equation, computing the inversion has complexity O(n3 ). So if we have a very large number of features, the normal equation will
be slow. In practice, when n exceeds 10,000 it might be a good time to go from a normal solution to an iterative process.

Normal Equation Noninvertibility

When implementing the normal equation in octave we want to use the 'pinv' function rather than 'inv.'

X T X may be noninvertible. The common causes are:

Redundant features, where two features are very closely related (i.e. they are linearly dependent)

Too many features (e.g. m ≤ n). In this case, delete some features or use "regularization" (to be explained in a later lesson).

Solutions to the above problems include deleting a feature that is linearly dependent with another or deleting one or more features when there are
too many features.

ML:Octave Tutorial

Basic Operations

https://www.coursera.org/learn/machine-learning/resources/QQx8l 4/680
07/01/2021 Machine Learning - Home | Coursera

1 %
%% Change Octave prompt  
2 PS1('>> ');
3 %% Change working directory in windows example:
4 cd 'c:/path/to/desired/directory name'
5 %% Note that it uses normal slashes and does not use escape characters for the empty spaces.
6
7 %% elementary operations
8 5+6
9 3-2
10 5*8
11 1/2
12 2^6
13 1 == 2 % false
14 1 ~= 2 % true.  note, not "!="
15 1 && 0
16 1 || 0
17 xor(1,0)
18
19
20 %% variable assignment
21 a = 3; % semicolon suppresses output
22 b = 'hi';
23 c = 3>=1;
24
25 % Displaying them:
26 a = pi
27 disp(a)
28 disp(sprintf('2 decimals: %0.2f', a))
29 disp(sprintf('6 decimals: %0.6f', a))
30 format long
31 a
32 format short
33 a
34
35
36 %%  vectors and matrices
37 A = [1 2; 3 4; 5 6]
38
39 v = [1 2 3]
40 v = [1; 2; 3]

Moving Data Around

Data files used in this section: featuresX.dat, priceY.dat

https://www.coursera.org/learn/machine-learning/resources/QQx8l 5/680
07/01/2021 Machine Learning - Home | Coursera

1 %
%% dimensions
2 sz = size(A) % 1x2 matrix: [(number of rows) (number of columns)]
3 size(A,1) % number of rows
4 size(A,2) % number of cols
5 length(v) % size of longest dimension
6
7
8 %% loading data
9 pwd   % show current directory (current path)
10 cd 'C:\Users\ang\Octave files'  % change directory 
11 ls    % list files in current directory 
12 load q1y.dat   % alternatively, load('q1y.dat')
13 load q1x.dat
14 who   % list variables in workspace
15 whos  % list variables in workspace (detailed view) 
16 clear q1y      % clear command without any args clears all vars
17 v = q1x(1:10); % first 10 elements of q1x (counts down the columns)
18 save hello.mat v;  % save variable v into file hello.mat
19 save hello.txt v -ascii; % save as ascii
20 % fopen, fread, fprintf, fscanf also work  [[not needed in class]]
21
22 %% indexing
23 A(3,2)  % indexing is (row,col)
24 A(2,:)  % get the 2nd row. 
25        % ":" means every element along that dimension
26 A(:,2)  % get the 2nd col
27 A([1 3],:) % print all  the elements of rows 1 and 3
28
29 A(:,2) = [10; 11; 12]     % change second column
30 A = [A, [100; 101; 102]]; % append column vec
31 A(:) % Select all elements as a column vector.
32
33 % Putting data together 
34 A = [1 2; 3 4; 5 6]
35 B = [11 12; 13 14; 15 16] % same dims as A
36 C = [A B]  % concatenating A and B matrices side by side
37 C = [A, B] % concatenating A and B matrices side by side
38 C = [A; B] % Concatenating A and B top and bottom
39

Computing on Data

https://www.coursera.org/learn/machine-learning/resources/QQx8l 6/680
07/01/2021 Machine Learning - Home | Coursera

1 %
%% initialize variables
2 A = [1 2;3 4;5 6]
3 B = [11 12;13 14;15 16]
4 C = [1 1;2 2]
5 v = [1;2;3]
6
7 %% matrix operations
8 A * C  % matrix multiplication
9 A .* B % element-wise multiplication
10 % A .* C  or A * B gives error - wrong dimensions
11 A .^ 2 % element-wise square of each element in A
12 1./v   % element-wise reciprocal
13 log(v)  % functions like this operate element-wise on vecs or matrices 
14 exp(v)
15 abs(v)
16
17 -v  % -1*v
18
19 v + ones(length(v), 1)  
20 % v + 1  % same
21
22 A'  % matrix transpose
23
24 %% misc useful functions
25
26 % max  (or min)
27 a = [1 15 2 0.5]
28 val = max(a)
29 [val,ind] = max(a) % val -  maximum element of the vector a and index - index value where maximum occur
30 val = max(A) % if A is matrix, returns max from each column
31
32 % compare values in a matrix & find
33 a < 3 % checks which values in a are less than 3
34 find(a < 3) % gives location of elements less than 3
35 A = magic(3) % generates a magic matrix - not much used in ML algorithms
36 [r,c] = find(A>=7)  % row, column indices for values matching comparison
37
38 % sum, prod
39 sum(a)
40 prod(a)

Plotting Data
1 %
%% plotting
2 t = [0:0.01:0.98];
3 y1 = sin(2*pi*4*t); 
4 plot(t,y1);
5 y2 = cos(2*pi*4*t);
6 hold on;  % "hold off" to turn off
7 plot(t,y2,'r');
8 xlabel('time');
9 ylabel('value');
10 legend('sin','cos');
11 title('my plot');
12 print -dpng 'myPlot.png'
13 close;           % or,  "close all" to close all figs
14 figure(1); plot(t, y1);
15 figure(2); plot(t, y2);
16 figure(2), clf;  % can specify the figure number
17 subplot(1,2,1);  % Divide plot into 1x2 grid, access 1st element
18 plot(t,y1);
19 subplot(1,2,2);  % Divide plot into 1x2 grid, access 2nd element
20 plot(t,y2);
21 axis([0.5 1 -1 1]);  % change axis scale
22
23 %% display a matrix (or image) 
24 figure;
25 imagesc(magic(15)), colorbar, colormap gray;
26 % comma-chaining function calls.  
27 a=1,b=2,c=3
28 a=1;b=2;c=3;
29

Control statements: for, while, if statements

https://www.coursera.org/learn/machine-learning/resources/QQx8l 7/680
07/01/2021 Machine Learning - Home | Coursera

1 v
v = zeros(10,1);
2 for i=1:10, 
3    v(i) = 2^i;
4 end;
5 % Can also use "break" and "continue" inside for and while loops to control execution.
6
7 i = 1;
8 while i <= 5,
9  v(i) = 100; 
10  i = i+1;
11 end
12
13 i = 1;
14 while true, 
15  v(i) = 999; 
16  i = i+1;
17  if i == 6,
18    break;
19  end;
20 end
21
22 if v(1)==1,
23  disp('The value is one!');
24 elseif v(1)==2,
25  disp('The value is two!');
26 else
27  disp('The value is not one or two!');
28 end
29

Functions
To create a function, type the function code in a text editor (e.g. gedit or notepad), and save the file as "functionName.m"

Example function:

1 function y = squareThisNumber(x)
f
2
3 y = x^2;
4

To call the function in Octave, do either:

1) Navigate to the directory of the functionName.m file and call the function:

1    % Navigate to directory:
2    cd /path/to/function
3
4    % Call the function:
5    functionName(args)
6

2) Add the directory of the function to the load path and save it:You should not use addpath/savepath for any of the assignments in this
course. Instead use 'cd' to change the current working directory. Watch the video on submitting assignments in week 2 for instructions.

1    % To add the path for the current session of Octave:
2    addpath('/path/to/function/')
3
4    % To remember the path for future sessions of Octave, after executing addpath above, also do:
5    savepath
6

Octave's functions can return more than one value:

https://www.coursera.org/learn/machine-learning/resources/QQx8l 8/680
07/01/2021 Machine Learning - Home | Coursera

1    function [y1, y2] = squareandCubeThisNo(x)
2    y1 = x^2
3    y2 = x^3
4

Call the above function this way:

1    [a,b] = squareandCubeThisNo(x)
2

Vectorization
Vectorization is the process of taking code that relies on loops and converting it into matrix operations. It is more efficient, more elegant, and
more concise.

As an example, let's compute our prediction from a hypothesis. Theta is the vector of fields for the hypothesis and x is a vector of variables.

With loops:

1 p
prediction = 0.0;
2 for j = 1:n+1,
3  prediction += theta(j) * x(j);
4 end;
5

With vectorization:

1 p
prediction = theta' * x;
2

If you recall the definition multiplying vectors, you'll see that this one operation does the element-wise multiplication and overall sum in a very
concise notation.

Working on and Submitting Programming Exercises

1. Download and extract the assignment's zip file.

2. Edit the proper file 'a.m', where a is the name of the exercise you're working on.

3. Run octave and cd to the assignment's extracted directory

4. Run the 'submit' function and enter the assignment number, your email, and a password (found on the top of the "Programming Exercises" page
on coursera)

Video Lecture Table of Contents

Basic Operations

1 0
0:00    Introduction
2 3:15    Elementary and Logical operations
3 5:12    Variables
4 7:38    Matrices
5 8:30    Vectors
6 11:53   Histograms
7 12:44   Identity matrices
8 13:14   Help command

Moving Data Around

https://www.coursera.org/learn/machine-learning/resources/QQx8l 9/680
07/01/2021 Machine Learning - Home | Coursera

1 0
0:24    The size command
2 1:39    The length command
3 2:18    File system commands
4 2:25    File handling
5 4:50    Who, whos, and clear
6 6:50    Saving data
7 8:35    Manipulating data
8 12:10   Unrolling a matrix
9 12:35   Examples
10 14:50   Summary

Computing on Data

1 0:00    Matrix operations
0
2 0:57    Element-wise operations
3 4:28    Min and max
4 5:10    Element-wise comparisons
5 5:43    The find command
6 6:00    Various commands and operations

Plotting data

1 0
0:00    Introduction
2 0:54    Basic plotting
3 2:04    Superimposing plots and colors
4 3:15    Saving a plot to an image
5 4:19    Clearing a plot and multiple figures
6 4:59    Subplots
7 6:15    The axis command
8 6:39    Color square plots
9 8:35    Wrapping up

Control statements

1 0
0:10    For loops
2 1:33    While loops
3 3:35    If statements
4 4:54    Functions
5 6:15    Search paths
6 7:40    Multiple return values
7 8:59    Cost function example (machine learning)
8 12:24   Summary
9

Vectorization

1 0
0:00    Why vectorize?
2 1:30    Example
3 4:22    C++ example
4 5:40    Vectorization applied to gradient descent
5 9:45    Python

https://www.coursera.org/learn/machine-learning/resources/QQx8l 10/680