티스토리 뷰

머신러닝

MSE, cost function

★ ☆ 2021. 1. 10. 11:25

 

J(θ0,θ1)=12mi=1m(y^iyi)2=12mi=1m(hθ(xi)yi)2

 

 

Hypothesis Function의 정확도를 cost function으로 구할 수 있다. 대개, cost function 값이 낮을 수 록 정확도가 높다. Cost function의 형태에 따라 달라질 수 있다.

 

Mean Squared Error는 대표적인 cost function 중의 하나이다.

hθ(xi)는 hypothesis function의 결과 값이며, yi는 실제 값이다.

 

 

*이해되지 않는 설명

The mean is halved (1/2) as a convenience for the computation of the gradient descent, as the derivative term of the square function will cancel out the (1/2) term.

Gradient descent 계산할 때 도움이 되기 위해서 mean을 절반으로 나눈다. 제곱 함수를 미분하면 1/2는 의미가 없어진다.

'머신러닝' 카테고리의 다른 글

Random Forest  (0) 2021.12.29
Skewed Class  (0) 2021.02.15
Batch gradient descent and Stochastic gradient descent  (0) 2021.01.10
Global minima and Local minima  (0) 2021.01.10
Categorical Variables  (0) 2021.01.08
최근에 올라온 글
최근에 달린 댓글
Total
Today
Yesterday
링크
«   2025/04   »
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30
글 보관함