English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Continual learning (CL) aims to acquire new knowledge while preservinginformation from previous experiences without forgetting. Though buffer-basedmethods (i.e., retaining samples from previous tasks) have achieved acceptableperformance, determining how to allocate the buffer remains a criticalchallenge. Most recent research focuses on refining these methods but oftenfails to sufficiently consider the varying influence of samples on the learningprocess, and frequently overlooks the complexity of the classes/concepts beinglearned. Generally, these methods do not directly take into account thecontribution of individual classes. However, our investigation indicates thatmore challenging classes necessitate preserving a larger number of samplescompared to less challenging ones. To address this issue, we propose a novelmethod and policy named 'Class-Adaptive Sampling Policy' (CASP), whichdynamically allocates storage space within the buffer. By utilizing concepts ofclass contribution and difficulty, CASP adaptively manages buffer space,allowing certain classes to occupy a larger portion of the buffer whilereducing storage for others. This approach significantly improves theefficiency of knowledge retention and utilization. CASP provides a versatilesolution to boost the performance and efficiency of CL. It meets the demand fordynamic buffer allocation, accommodating the varying contributions of differentclasses and their learning complexities over time.

Class-Adaptive Sampling Policy for Efficient Continual Learning