Regularized logistic regression code in matlab

 I'm trying my hand at regularized LR, simple with this formulas in matlab:

The cost function: 

 

J(theta) = 1/m*sum((-y_i)*log(h(x_i)-(1-y_i)*log(1-h(x_i))))+(lambda/2*m)*sum(theta_j)

The gradient: 

 

∂J(theta)/∂theta_0 = [(1/m)*(sum((h(x_i)-y_i)*x_j)] if j=0

∂j(theta)/∂theta_n = [(1/m)*(sum((h(x_i)-y_i)*x_j)]+(lambda/m)*(theta_j) if j>1

This is not matlab code is just the formula. 

So far I've done this: 

 

function [J, grad] = costFunctionReg(theta, X, y, lambda)

J = 0;
grad = zeros(size(theta));

temp_theta = [];

%cost function

%get the regularization term

for jj = 2:length(theta)

    temp_theta(jj) = theta(jj)^2;
end

theta_reg = lambda/(2*m)*sum(temp_theta);

temp_sum =[];

%for the sum in the cost function

for ii =1:m

   temp_sum(ii) = -y(ii)*log(sigmoid(theta'*X(ii,:)'))-(1-y(ii))*log(1-sigmoid(theta'*X(ii,:)'));

end

tempo = sum(temp_sum);

J = (1/m)*tempo+theta_reg;

%regulatization
%theta 0

reg_theta0 = 0;

for jj=1:m
 reg_theta0(jj) = (sigmoid(theta'*X(m,:)') -y(jj))*X(jj,1)
end    

reg_theta0 = (1/m)*sum(reg_theta0)

grad_temp(1) = reg_theta0

%for the rest of thetas

reg_theta  = [];
thetas_sum = 0;

for ii=2:size(theta)
    for kk =1:m
        reg_theta(kk) = (sigmoid(theta'*X(m,:)') - y(kk))*X(kk,ii)
    end
    thetas_sum(ii) = (1/m)*sum(reg_theta)+(lambda/m)*theta(ii)
    reg_theta = []
end

for i=1:size(theta)

    if i == 1
        grad(i) = grad_temp(i)
    else
        grad(i) = thetas_sum(i)
    end
end
end

And the cost function is giving correct results, but I have no idea why the gradient (one step) is not, the cost gives J = 0.6931 which is correct and the gradient grad = 0.3603 -0.1476 0.0320, which is not, the cost starts from 2 because the parameter theta(1) does not have to be regularized, any help? I guess there is something wrong with the code, but after 4 days I can't see it.Thanks  

NOTE:-


Matlabsolutions.com provide latest MatLab Homework Help,MatLab Assignment Help , Finance Assignment Help for students, engineers and researchers in Multiple Branches like ECE, EEE, CSE, Mechanical, Civil with 100% output.Matlab Code for B.E, B.Tech,M.E,M.Tech, Ph.D. Scholars with 100% privacy guaranteed. Get MATLAB projects with source code for your learning and research. 

Comments

Post a Comment

Popular posts from this blog

How to control the order of detected objects by regionprops in matlab?

  Answer The  regionprops  function in MATLAB returns a struct array containing various properties of the labeled regions in an image. The order of the objects detected by  regionprops  is determined by the order in which the objects appear in the labeled image. If you want to control the order of the detected objects, you can modify the labeling process of the image. One approach is to use a different labeling algorithm that generates the labels in the order you desire. For example, you can use the  bwlabel  function with the 'linear' option, which labels the objects in the order they appear from left to right and top to bottom in the image. Here's an example code snippet:    % Load binary image binaryImage = imread('example.png'); NOTE:- Matlabhelpers.com  provide latest  MatLab Homework Help , MatLab Assignment Help  ,  Finance Assignment Help  for students, engineers and researchers in Multiple Branches like ECE, ...
Read more

How is full convolution performed using MATLAB's conv2 function?

  Answer In MATLAB, the  conv2  function is used to perform 2D convolution between two matrices. The syntax of  conv2  is as follows:    C = conv2(A, B) Here,  A  and  B  are input matrices, and  C  is the output matrix after convolution. When  A  and  B  have the same size, the convolution is called "full convolution," which means that the output matrix  C  has the same size as the sum of the sizes of  A  and  B  minus 1. To perform full convolution using  conv2 , the input matrices  A  and  B  are first flipped both horizontally and vertically. Then, the flipped  B  matrix is slid over  A  matrix, and the dot product of the overlapped elements is computed and added up. This process is repeated for all possible positions of  B  on  A . The resulting matrix is the full convolution output  C . Here is an exampl...
Read more

Why do I get a "Too many input arguments" error when not passing any?

  I am working on some simple object-oriented code in MATLAB. I am trying to call one of my class methods with no input or output arguments in its definition. Function definition: function roll_dice Function call: obj.roll_dice; When this is executed, MATLAB says:    ??? Error using ==> roll_dice Too many input arguments. Error in ==> DiceSet>Diceset.Diceset at 11 obj.roll_dice; (etc...) Anyone have any ideas what could be causing it? Are there secret automatic arguments I'm unaware that I'm passing?   NOTE:- Matlabsolutions.com  provide latest  MatLab Homework Help, MatLab Assignment Help  ,  Finance Assignment Help  for students, engineers and researchers in Multiple Branches like ECE, EEE, CSE, Mechanical, Civil with 100% output.Matlab Code for B.E, B.Tech,M.E,M.Tech, Ph.D. Scholars with 100% privacy guaranteed. Get MATLAB projects with source code for your learning and research.     Answers:  When you ma...
Read more