728x90 반응형 early stopping2 Double Descent: new approach of bias-variance trade-off Double Descent: Modern approach of bias-variance trade-offclassic ML의 관점에서, bias-variance trade-off는 model complexity와 관련있다.이전 포스트를 참고해도 좋다.https://trivia-starage.tistory.com/238 Overfitting을 막는 방법들 (regularization, cross-validation, early stopping)Overfitting and Regularization ※ The blog post is based on lecture materials from Xavier Bresson, a professor at the National University of Singapore.. 2024. 3. 3. Overfitting을 막는 방법들 (regularization, cross-validation, early stopping) Overfitting and Regularization※ The blog post is based on lecture materials from Xavier Bresson, a professor at the National University of Singapore. The lecture materials can be found on the professor's LinkedIn. You can also found it at [1]. Under-fitting and over-fittingUnderfittinglearner가 충분한 표현력을 가지지 못함.training set에서 error를 생성.training/testing error 모두 높다.방지 방법: learner의 expressivity(또는 c.. 2024. 3. 2. 이전 1 다음 728x90 반응형
