Year: 2024
Type: preprint
Abstract: Our paper addresses the problem of models struggling to learn diverse features, due to either forgetting previously learned features or failing to learn new ones. To overcome this problem, we introduce Diverse Feature Learning (DFL), a method that combines an important feature preservation algorithm with a new feature learning algorithm. Specifically, for preserving important features, we utilize self-distillation in ensemble models by selecting the meaningful model weights observed during training. For learning new features, we employ reset that involves periodically re-initializing part of the model. As a result, through experiments with various models on the image classification, we have identified the potential for synergistic effects between self-distillation and reset.
Source: arXiv (Cornell University)
Author: Park, Sejik
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4393398983