언젠가 알맞은 카테고리로 (2) 썸네일형 리스트형 Notes for Coursera - Machine Learning Gradient Descent Feature Scaling + mean normalisation: make sure features are on a similar scale and have a mean of 0, xi = xi - ui / si, where ui is average and si is range. Debugging Try plotting min(J(theta)) vs #iterations as we go, to see if the value is decreasing properly try tripling learning rate Normal Equation: inv(trans_X * X) * trans_X * Y / no need for feature scaling Compared to g.. 유용한 링크 matplotlib - Matplotlib 2 y values against each X (xspdf.com) 이전 1 다음