公开(公告)号：US20240354593A1

公开(公告)日：2024-10-24

申请号：US18355672

申请日：2023-07-20

Applicant: Google LLC

Inventor： Rodolphe René Willy Jenatton ,  Mark Patrick Collier ,  Effrosyni Kokiopoulou ,  Basil Mustafa ,  Neil Matthew Tinmouth Houlsby ,  Jesse Berent

IPC: G06N3/0985 ,  G06N3/048

CPC classification number: G06N3/0985 ,  G06N3/048

Abstract: HET classifiers, which learn a multivariate Gaussian distribution over prediction logits, perform well on image classification problems with hundreds to thousands of classes. However, compared to standard classifiers (e.g., deterministic (DET) classifiers), they introduce extra parameters that scale linearly with the number of classes. This makes them infeasible to apply to larger-scale problems. In addition, HET classifiers introduce a temperature hyperparameter, which is ordinarily tuned. HET classifiers are disclosed, where the parameter count (when compared to a DET classifier) scales independently of the number of classes. In large-scale settings of the embodiments, the need to tune the temperature hyperparameter is removed, by directly learning it on the training data.