LGFeb 15, 2021

Online hyperparameter optimization by real-time recurrent learning

arXiv:2102.07813v29 citations
Originality Incremental advance
AI Analysis

This addresses the problem of efficiently optimizing hyperparameters in dynamic scenarios like life-long learning for machine learning practitioners, though it is incremental as it adapts existing RNN learning methods.

The paper tackles the computational intensity and lack of adaptability in conventional hyperparameter optimization by proposing an online algorithm that simultaneously tunes hyperparameters and network parameters, achieving better generalization performance at a fraction of the wallclock time compared to standard methods.

Conventional hyperparameter optimization methods are computationally intensive and hard to generalize to scenarios that require dynamically adapting hyperparameters, such as life-long learning. Here, we propose an online hyperparameter optimization algorithm that is asymptotically exact and computationally tractable, both theoretically and practically. Our framework takes advantage of the analogy between hyperparameter optimization and parameter learning in recurrent neural networks (RNNs). It adapts a well-studied family of online learning algorithms for RNNs to tune hyperparameters and network parameters simultaneously, without repeatedly rolling out iterative optimization. This procedure yields systematically better generalization performance compared to standard methods, at a fraction of wallclock time.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes