everpeace / ml-class-assignments Goto Github PK

*[cost, grad] = costFunction(initial_theta, X, y); error: costFunction: operator : nonconformant arguments (op1 is 1x100, op2 is 1x100) error: called from costFunction at line 16 column 3

sigma3 = a3.-Y;

There is no need to use "." in subtraction
It must be -->

sigme=a3-Y

Should be
J = sum(sum((diff.^2).*(R==1)))/2;
in lieu of
J = sum((diff.^2)(R==1))/2;

Hi,
I am wondering how to derive the logic for prediction

p = sigmoid(X*theta)>=0.5;

is there any standard formula?

Hello, why is the grid range defined as such?
https://github.com/everpeace/ml-class-assignments/blob/master/ex2.Logistic_Regression/mlclass-ex2/plotDecisionBoundary.m#L29-L31

J = (1/2) * sum(sum((R .* ((Theta * X')' ) - Y).^2)) + (lambda/2) * sum(sum(Theta.^2)) + (lambda/2) * sum(sum(X.^2));

X_grad = (R .* ((Theta * X')' ) - Y) * Theta + (lambda * X );
Theta_grad = (R .* ((Theta * X')' ) - Y)' * X + (lambda * Theta);

Pardon me for asking a trivial question.

Can you help me on calculating the decision boundary line. How it has been "plot_y = (-1./theta(3)).*(theta(2).*plot_x + theta(1));"

Guys request your help!!

Instead of using the code mentioned in code section :

p=sigmoid(X*theta)>=0.5;

which is running perfectly fine. Im using my logic :

z=sigmoid(X*theta);
if z>=0.5:
p=1;
else
p=0;
endif

The above code is not running.

Please let me know the difference and is there any error in my logic.

Hello! I think that line should be:

sigma3=a3-Y;

Instead of, ".-" is not supported in matlab: