일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
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 |
- 맥북원스토어
- ios원스토어
- 백준
- 선대
- CS231nSVM
- CNN구조정리
- 맥실리콘
- Linear algebra
- CS231nAssignment1
- CS231ntwolayerneuralnet
- CS231nAssignments
- ㅐㅕ세ㅕㅅ
- pycharmerror
- BAEKJOON
- arm칩에안드로이드
- professor strang
- BOJ
- 아이폰원스토어
- CNNarchitecture
- 백준알고리즘
- Algorithm
- RegionProposalNetworks
- Gilbert Strang
- MIT
- adversarialattackonmonoculardepthestimation
- monoculardepthestimation
- MacOS
- 선형대수학
- gpumemory
- CS231n
- Today
- Total
목록CS231nSVM (2)
개발로 하는 개발

Softmax Another popular classifier (like SVM) Generalized version of binary Logistic Regression classifier - Softmax function : - Loss function : cross-entropy function - Numerical Stability : exponential -> very large number -> normalize values - SVM vs Softmax Same score function Wx = b, different loss function. Softmax : Calculate probabilities for each classes. Easier to interpret. Softmax.p..

KNN - space inefficient : have to remember all the data in the training set - classifying is expensive : must calculate all the distances to all of the training set -> Use SVM SVM Linear Classification - Score function, Loss function 사용 : minimize the loss function with respect to the parameters of the score function. CIFAR-10 we have a training set of N = 50,000 images, each with D = 32 x 32 x ..