I get an error when I try to run my logistic regression.

This is my gradient function for logistic regression, but when I try to run it, I get an error that says "the left and right sides have a different number of elements". It refers to line 8. Any suggestions?

function [theta,J] = gradient(X_training, Y_training, alpha, theta, iterations)

J = 0; % Initial cost

m = length(Y_training);

for j = 1:iterations

h = sigmoid(X_training * theta);

error = h - Y_training;

for i = 1:length(theta)

line 8 -----> theta(i) = theta(i) - (alpha/m) * sum(error .* X_training(:,i)); % Updating thetas

end

J(j) = cost_function(X_training,Y_training,theta); % Cost of each iteration

end

end

Answer 1

[Torsten]

theta is not initialized - thus the theta(i) at the right-hand side of the assignment

theta(i) = theta(i) - (alpha/m) * sum(error .* X_training(:,i));   % Updating thetas 

will throw an error.

Further,

 sum(error .* X_training(:,i))

doesn't seem to be a scalar value, but a vector.

Maybe

 sum((error.') .* X_training(:,i))

has to be used instead.

Answer 2

[David Hill]

function [J, grad] = costFunction(theta, X, y)
m = length(y);
h=sigmoid(X*theta);
J=sum((-y.*log(h)-(1-y).*log(1-h)))/m;
grad=X'*(h-y)/m;
end

function g = sigmoid(z)
g=1./(1+exp(-z));
end

[Seyed Navid Shoaiby]

This code works, but I want to keep track of "J" in each iteration. I need a "for loop" to do that.