Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

안녕하세요 질문이 있어 남깁니다 #1

Open
Tajyu opened this issue Aug 3, 2021 · 1 comment
Open

안녕하세요 질문이 있어 남깁니다 #1

Tajyu opened this issue Aug 3, 2021 · 1 comment

Comments

@Tajyu
Copy link

Tajyu commented Aug 3, 2021

3가지의 클래스로 분류를 하고싶은데 하나의 클래스만 데이터 수가 1/11 또는 1/10로 다른 클래스들과 차이가나는데 학습전에 데이터 불균형을 어떻게 잡는게 좋을까요...?ㅠㅠ

@lovedlim
Copy link
Owner

lovedlim commented Aug 3, 2021

안녕하세요:) 말씀하신 것처럼 불균형이 심할 때 일반적으로 SMOTE, GAN, 언더/오버샘플링 기법등이 있습니다. 하지만 실제 적용했을 때 효과가 있을지는 해봐야 알 것 같아요! 또한 딥러닝에서는 weight balancing 값을 줄 수 있습니다. weight balancing 을 통해서 (데이터가 작은)특정 클래스에 대해 loss값을 크게 만들 수 있어요!
참고 자료: 텐서플로 튜토리얼 "클래스 가중치로 모델 교육" 부분 참고하세요!
https://www.tensorflow.org/tutorials/structured_data/imbalanced_data?hl=ko

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants