Unsupervised Continual Learning via Self-Adaptive Deep Clustering Approach

TL;DR

This paper introduces KIERA, an unsupervised continual learning method that uses self-adaptive deep clustering with elastic networks and centroid-based experience replay to handle changing environments without labeled data.

Contribution

KIERA is a novel unsupervised continual learning framework that does not require task labels and employs elastic networks and clustering for knowledge retention.

Findings

KIERA achieves competitive performance on continual learning benchmarks.

It effectively mitigates catastrophic forgetting without labeled data.

The approach is task-agnostic and adaptable to changing environments.

Abstract

Unsupervised continual learning remains a relatively uncharted territory in the existing literature because the vast majority of existing works call for unlimited access of ground truth incurring expensive labelling cost. Another issue lies in the problem of task boundaries and task IDs which must be known for model's updates or model's predictions hindering feasibility for real-time deployment. Knowledge Retention in Self-Adaptive Deep Continual Learner, (KIERA), is proposed in this paper. KIERA is developed from the notion of flexible deep clustering approach possessing an elastic network structure to cope with changing environments in the timely manner. The centroid-based experience replay is put forward to overcome the catastrophic forgetting problem. KIERA does not exploit any labelled samples for model updates while featuring a task-agnostic merit. The advantage of KIERA has been numerically validated in popular continual learning problems where it shows highly competitive performance compared to state-of-the art approaches. Our implementation is available in \textit{\url{https://github.com/ContinualAL/KIERA}}.