Class-Adaptive Sampling Policy for Efficient Continual Learning

TL;DR

This paper introduces CASP, a class-adaptive sampling policy for continual learning that dynamically allocates buffer space based on class difficulty and contribution, improving knowledge retention and efficiency.

Contribution

The paper proposes a novel buffer management method, CASP, that adaptively allocates storage based on class difficulty and contribution, addressing limitations of existing buffer-based continual learning methods.

Findings

CASP improves knowledge retention efficiency in continual learning.

CASP dynamically allocates buffer space based on class difficulty.

Experimental results show enhanced performance over traditional methods.

Abstract

Continual learning (CL) aims to acquire new knowledge while preserving information from previous experiences without forgetting. Though buffer-based methods (i.e., retaining samples from previous tasks) have achieved acceptable performance, determining how to allocate the buffer remains a critical challenge. Most recent research focuses on refining these methods but often fails to sufficiently consider the varying influence of samples on the learning process, and frequently overlooks the complexity of the classes/concepts being learned. Generally, these methods do not directly take into account the contribution of individual classes. However, our investigation indicates that more challenging classes necessitate preserving a larger number of samples compared to less challenging ones. To address this issue, we propose a novel method and policy named 'Class-Adaptive Sampling Policy' (CASP), which dynamically allocates storage space within the buffer. By utilizing concepts of class contribution and difficulty, CASP adaptively manages buffer space, allowing certain classes to occupy a larger portion of the buffer while reducing storage for others. This approach significantly improves the efficiency of knowledge retention and utilization. CASP provides a versatile solution to boost the performance and efficiency of CL. It meets the demand for dynamic buffer allocation, accommodating the varying contributions of different classes and their learning complexities over time.