Exercise 2 Contour Plot for Lambda values isn't getting accuracy about ml-coursera-python-assignments HOT 5 OPEN

This matter also troubles me.
no matter the lambda value and max iterators are, the accuracy is always around 66%

FYI, I changed the method used for the optimisation to 'BFGS' and I get the answer required. Maybe someone who is more an expert in numerical computing can answer this better as it seems the choice of optimisation method is what is causing the problem.

from ml-coursera-python-assignments.

This matter also troubles me.
no matter the lambda value and max iterators are, the accuracy is always around 66%

from ml-coursera-python-assignments.

This matter also troubles me.
no matter the lambda value and max iterators are, the accuracy is always around 66%

I am getting the exact same values as you, I don't entirely understand what is going on since I've submitted to the grader all of my previous functions and got all the expected values that were required. Did you find the solution this problem?

from ml-coursera-python-assignments.

I have the same problem using TNC method, but the code works if I change to CG method:

res = optimize.minimize(costFunctionReg,
initial_theta,
(X, y, lambda_),
jac=True,
method='CG',
options=options)

Anyone can please help explain?...

from ml-coursera-python-assignments.

I am getting 60.2% with TNC as well... Changed to CG and got Train Accuracy: 83.1 %..

Someone please help explain...
THX..

from ml-coursera-python-assignments.