I'm trying to implement Logistic Regression with Gradient Descent. But the w value is getting apparently random values when I run the code. I know I'm doing something wrong, for instance I don't know the proper value for ALPHA and an initial value for MAX_ANGLE. So I hope you guys help me to set this up properly.
What am I doing wrong? I know I'm doing something wrong, for instance I don't know the proper value for ALPHA and a proper initial value for "w". So I hope you guys help me to set this up properly.
I never heard of "Logistic Regression with Gradient Descent" before, so I spent a minute or two reading up about it.
I got far enough to realize that your code was meant to implement stochastic gradient descent, but really all that I can say is that it would be very easy to make an error while converting those calculations to Java code.
However it seems like you don't know what output is supposed to come out of that code, so I suppose there's a possibility that it's actually correct? Given that you have random inputs, it maybe isn't surprising that the outputs look "random"?