본문 바로가기
ML | 데이터과학/머신러닝

Linear regrssion, cost func. , Logistic

by 노아론 2018. 1. 26.
linear regression, cost, logistic

Linear Regression 정리

Hypothesis

H(x)=Wx+bH(x)=W*x+b
H(x1,x2,x3)=w1x1+w2x+w3x3+bH(x1,x2,x3)=w_{1}x_{1}+w_{2}x+w_{3}x_{3}+b

실제 구현시 H(x)=XH
(매트릭스를 사용한다)
bias는 간략히 하기위해 생략

Cost Function

cost(W,b)=1mi=1m(H(x)iyi)2)cost(W,b) = \frac{1}{m}\sum_{i=1}^{m}(H(x)^i-y^i)^2)
cost(W,b)=1mi=1m(H(x1,x2,x3)iyi)2cost(W,b) = \frac{1}{m}\sum_{i=1}^{m}(H(x_{1},x_{2},x_{3})^i-y^i)^2

Logistic hypothesis

H(x)=11+e(WTX)H(x)=\frac{1}{1+e^(-W^T X)}

댓글