Genetically-regulated Neuromodulation Facilitates Multi-Task Reinforcement Learning - Archive ouverte HAL Access content directly
Conference Papers Year : 2015

Genetically-regulated Neuromodulation Facilitates Multi-Task Reinforcement Learning

Abstract

In this paper, we use a gene regulatory network (GRN) to regulate a reinforcement learning controller, the State- Action-Reward-State-Action (SARSA) algorithm. The GRN serves as a neuromodulator of SARSA's learning parame- ters: learning rate, discount factor, and memory depth. We have optimized GRNs with an evolutionary algorithm to regulate these parameters on specific problems but with no knowledge of problem structure. We show that genetically- regulated neuromodulation (GRNM) performs comparably or better than SARSA with fixed parameters. We then ex- tend the GRNM SARSA algorithm to multi-task problem generalization, and show that GRNs optimized on multi- ple problem domains can generalize to previously unknown problems with no further optimization.
Fichier principal
Vignette du fichier
cussat_15413.pdf (1.38 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04077782 , version 1 (21-04-2023)

Identifiers

Cite

Sylvain Cussat-Blanc, Kyle Harrington. Genetically-regulated Neuromodulation Facilitates Multi-Task Reinforcement Learning. Genetic and Evolutionary Computation Conference (GECCO 2015), Jul 2015, Madrid, Spain. pp.551-558, ⟨10.1145/2739480.2754730⟩. ⟨hal-04077782⟩
17 View
12 Download

Altmetric

Share

Gmail Facebook X LinkedIn More