반응형
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
- 머신러닝
- java역사
- C언어
- 결합전문기관
- vscode
- KNeighborsClassifier
- broscoding
- pycharm
- bccard
- discrete_scatter
- Keras
- mglearn
- web 개발
- 데이터전문기관
- CES 2O21 참여
- CES 2O21 참가
- 자료구조
- paragraph
- tensorflow
- 대이터
- web 용어
- classification
- html
- web 사진
- postorder
- inorder
- 재귀함수
- 웹 용어
- web
- cudnn
Archives
- Today
- Total
bro's coding
recommendation system.word2vec 본문
반응형
2003년 : Bengio -> 2013년 : Mikolov(google)
정확도 향상 / 성능 향상(Skip-gram) : acc가 약 2배 좋아짐
빠른 훈련(최적화 기술) : train 7일 -> 2일
- 장점
사용이 심플하다.
라벨링 되지 않은 데이터로 이용 가능하다.
사용의 편리성에 비해 엄청난 정확도를 보여준다.
- 단점
왜 이렇게 결과가 나오는지 명확한 해석이 어렵다.(머신러닝 대표 특징)
의도적인 성능 향상을 위해서는 많은 삽질이 필요하다.
반응형
'[INTURNSHIP] > HuminTec' 카테고리의 다른 글
word2Vec (0) | 2021.01.25 |
---|---|
NMP (0) | 2021.01.25 |
pyQt5.reference (0) | 2021.01.21 |
Nvivo (0) | 2021.01.11 |
recommendation system.word2vec.NNLM(neural net language model) (0) | 2021.01.06 |
recommendation system.reference (0) | 2021.01.05 |
recommendation system.library.surprise.info.install (0) | 2021.01.04 |
recommendation system.classification.feature (0) | 2021.01.04 |
Comments