Incremental Task Learning with Incremental Rank Updates

TL;DR

This paper introduces a novel incremental task learning method using low-rank matrix updates, improving accuracy and memory efficiency over existing approaches by representing network weights as combinations of rank-1 matrices.

Contribution

The paper proposes a new low-rank factorization framework for incremental task learning that updates network weights with rank-1 matrices and a selector vector, outperforming current state-of-the-art methods.

Findings

Achieves higher accuracy than existing methods.

Reduces catastrophic forgetting effectively.

Offers better memory efficiency.

Abstract

Incremental Task learning (ITL) is a category of continual learning that seeks to train a single network for multiple tasks (one after another), where training data for each task is only available during the training of that task. Neural networks tend to forget older tasks when they are trained for the newer tasks; this property is often known as catastrophic forgetting. To address this issue, ITL methods use episodic memory, parameter regularization, masking and pruning, or extensible network structures. In this paper, we propose a new incremental task learning framework based on low-rank factorization. In particular, we represent the network weights for each layer as a linear combination of several rank-1 matrices. To update the network for a new task, we learn a rank-1 (or low-rank) matrix and add that to the weights of every layer. We also introduce an additional selector vector that assigns different weights to the low-rank matrices learned for the previous tasks. We show that our approach performs better than the current state-of-the-art methods in terms of accuracy and forgetting. Our method also offers better memory efficiency compared to episodic memory- and mask-based approaches. Our code will be available at https://github.com/CSIPlab/task-increment-rank-update.git