On the consistency of hyper-parameter selection in value-based deep reinforcement learning

TL;DR

This paper empirically investigates the reliability of hyper-parameter choices in value-based deep reinforcement learning, introducing a new score to measure their consistency and identifying which hyper-parameters are most critical and stable.

Contribution

It provides an extensive empirical analysis of hyper-parameter consistency in deep RL and introduces a new metric for evaluating hyper-parameter reliability.

Findings

Certain hyper-parameters are consistently critical across different settings.

Some hyper-parameters show high stability and do not require frequent retuning.

The new score effectively quantifies hyper-parameter reliability.

Abstract

Deep reinforcement learning (deep RL) has achieved tremendous success on various domains through a combination of algorithmic design and careful selection of hyper-parameters. Algorithmic improvements are often the result of iterative enhancements built upon prior approaches, while hyper-parameter choices are typically inherited from previous methods or fine-tuned specifically for the proposed technique. Despite their crucial impact on performance, hyper-parameter choices are frequently overshadowed by algorithmic advancements. This paper conducts an extensive empirical study focusing on the reliability of hyper-parameter selection for value-based deep reinforcement learning agents, including the introduction of a new score to quantify the consistency and reliability of various hyper-parameters. Our findings not only help establish which hyper-parameters are most critical to tune, but also help clarify which tunings remain consistent across different training regimes.