Notice
Recent Posts
Recent Comments
Link
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 |
Tags
- NPM
- nodejs
- 런데이
- 단당류
- 화학구조
- 리액트
- 키체인
- 텐서플로우
- 파일이름
- 달리는사이
- AI 전시
- 러닝
- 달리기
- tensorflow
- npx
- AI엑스포
- 2022 AI EXPO
- ipad
- 딥러닝
- 키체인초기화
- create-react-app
- 따릉이
- AI 회사
- react
- 6일차
- Nike
- 국제인공지능대전
- 연구
- NRC
- Running
Archives
- Today
- Total
Tomato's BLOG
Batch Normalization layer의 위치 및 Normalization 종류 본문
Batch Normalization layer 위치
Activation function(AF)에 따른 Batch normalization layer(BN)의 위치
- s자형(sigmoid, tanh 등)은 AF 뒤에 BN
- 선형(ReLU 등)은 AF 앞에 BN
https://machinelearningmastery.com/batch-normalization-for-training-of-deep-neural-networks/
A Gentle Introduction to Batch Normalization for Deep Neural Networks
Training deep neural networks with tens of layers is challenging as they can be sensitive to the initial random weights […]
machinelearningmastery.com
Normalization의 종류
https://sanghyu.tistory.com/32?category=1120072
[PyTorch] 시계열 데이터를 위한 다양한 Normalization 기법(BatchNorm1d, GroupNorm 사용법)
Normalization Neural network의 깊이가 점점 깊어질수록 학습이 안정적으로 되지 않는 문제가 발생한다. 이런 학습의 불안정화의 원인으로 'internal covariance shift'가 언급되었다. 이는 딥러닝에서 각 layer.
sanghyu.tistory.com
'개발 > 머신러닝, 딥러닝' 카테고리의 다른 글
AutoML 참고 사이트 정리 (0) | 2021.04.22 |
---|---|
[Pandas] DataFrame 데이터 확인 (0) | 2021.03.11 |
[딥러닝] 텐서플로우 데이터셋(tensorflow_datasets) 활용 (0) | 2021.02.04 |
Comments