Saved in:
Bibliographic Details
Main Author: Yang, Fan
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.16517
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909300408975360
author Yang, Fan
author_facet Yang, Fan
contents Variational continual learning (VCL) is a turn-key learning algorithm that has state-of-the-art performance among the best continual learning models. In our work, we explore an extension of the generalized variational continual learning (GVCL) model, named AutoVCL, which combines task heuristics for informed learning and model optimization. We demonstrate that our model outperforms the standard GVCL with fixed hyperparameters, benefiting from the automatic adjustment of the hyperparameter based on the difficulty and similarity of the incoming task compared to the previous tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2408_16517
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Adaptive Variational Continual Learning via Task-Heuristic Modelling
Yang, Fan
Machine Learning
Artificial Intelligence
Variational continual learning (VCL) is a turn-key learning algorithm that has state-of-the-art performance among the best continual learning models. In our work, we explore an extension of the generalized variational continual learning (GVCL) model, named AutoVCL, which combines task heuristics for informed learning and model optimization. We demonstrate that our model outperforms the standard GVCL with fixed hyperparameters, benefiting from the automatic adjustment of the hyperparameter based on the difficulty and similarity of the incoming task compared to the previous tasks.
title Adaptive Variational Continual Learning via Task-Heuristic Modelling
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2408.16517