Nonton Film: 1288469 can overfitted deep neural networks in adversarial training generalize an approximation viewpoint 2024 Sub Indo

Can overfitted deep neural networks in adversarial training generalize? – An approximation viewpoint (2024)

Overview: In this talk I will discuss whether overfitted DNNs in adversarial training can generalize from an approximation viewpoint. We prove by construction the existence of infinitely many adversarial training classifiers on over-parameterized DNNs that obtain arbitrarily small adversarial training error (overfitting) whereas achieving good robust generalization error under certain conditions concerning the data quality well separated and perturbation level. This construction is optimal and thus points out the fundamental limits of DNNs under adversarial training with statistical guarantees. Part of this talk comes from our recent work.

Genres: Documentary, Crime

Director: Fanghui Liu

Cast: Fanghui Liu, Engineering Research Building, Room 514, Yuchen Zeng

Original Language: EN

Original Title: Can overfitted deep neural networks in adversarial training generalize? – An approximation viewpoint

Budget: N/A

Revenue: N/A

MPAA Rating: PG

Keywords:

▶️ Watch Now