Directed Structural Adaptation to Overcome Statistical Conflicts and Enable Continual Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Erden, Zeki Doruk, Faltings, Boi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912145582587904
author Erden, Zeki Doruk
Faltings, Boi
author_facet Erden, Zeki Doruk
Faltings, Boi
contents Adaptive networks today rely on overparameterized fixed topologies that cannot break through the statistical conflicts they encounter in the data they are exposed to, and are prone to "catastrophic forgetting" as the network attempts to reuse the existing structures to learn new task. We propose a structural adaptation method, DIRAD, that can complexify as needed and in a directed manner without being limited by statistical conflicts within a dataset. We then extend this method and present the PREVAL framework, designed to prevent "catastrophic forgetting" in continual learning by detection of new data and assigning encountered data to suitable models adapted to process them, without needing task labels anywhere in the workflow. We show the reliability of the DIRAD in growing a network with high performance and orders-of-magnitude simpler than fixed topology networks; and demonstrate the proof-of-concept operation of PREVAL, in which continual adaptation to new tasks is observed while being able to detect and discern previously-encountered tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2412_04190
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Directed Structural Adaptation to Overcome Statistical Conflicts and Enable Continual Learning
Erden, Zeki Doruk
Faltings, Boi
Machine Learning
Artificial Intelligence
Adaptive networks today rely on overparameterized fixed topologies that cannot break through the statistical conflicts they encounter in the data they are exposed to, and are prone to "catastrophic forgetting" as the network attempts to reuse the existing structures to learn new task. We propose a structural adaptation method, DIRAD, that can complexify as needed and in a directed manner without being limited by statistical conflicts within a dataset. We then extend this method and present the PREVAL framework, designed to prevent "catastrophic forgetting" in continual learning by detection of new data and assigning encountered data to suitable models adapted to process them, without needing task labels anywhere in the workflow. We show the reliability of the DIRAD in growing a network with high performance and orders-of-magnitude simpler than fixed topology networks; and demonstrate the proof-of-concept operation of PREVAL, in which continual adaptation to new tasks is observed while being able to detect and discern previously-encountered tasks.
title Directed Structural Adaptation to Overcome Statistical Conflicts and Enable Continual Learning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2412.04190