RSAC: Regularized Subspace Approximation Classifier for Lightweight Continuous Learning
This addresses the need for lightweight continuous learning in constrained environments like edge computing, though it appears incremental as it builds on existing solutions.
The paper tackles the problem of excessive training time and memory usage in continuous learning by proposing RSAC, a novel training algorithm with feature reduction and regularization. Experiments show RSAC is more efficient and outperforms prior works across various settings.
Continuous learning seeks to perform the learning on the data that arrives from time to time. While prior works have demonstrated several possible solutions, these approaches require excessive training time as well as memory usage. This is impractical for applications where time and storage are constrained, such as edge computing. In this work, a novel training algorithm, regularized subspace approximation classifier (RSAC), is proposed to achieve lightweight continuous learning. RSAC contains a feature reduction module and classifier module with regularization. Extensive experiments show that RSAC is more efficient than prior continuous learning works and outperforms these works on various experimental settings.