coursera machine learning note

To be continued..

-----------------------------------------------------------------

Last updated: 2016.4.2.

SVM is a trick that allows the computer to deal with infinite features.

 

Optimization algorithm:
gradient descent

conjugate gradient

BFGS

L-BFGS

 

underfit, high bias

just right

overfit, high variance

 

convergence is
achieved (all training samples are classified correctly)