Year: 2003
Type: conference-paper
Abstract: Relational reinforcement learning (RRL) is a Q-learning technique which uses rst order regression techniques to generalize the Qfunction.Both the relational setting and the Q-learning context introduce a number of difculties which must be dealt with.In this paper we investigate a few dierent methods that do incremental relational instance based regression and can be used for RRL.This leads us to dierent approaches which limit both memory consumption and processing times.We implemented a number of these approaches and experimentally evaluated and compared them to each other and an existing RRL algorithm.These experiments show relational instance based regression to work well and to add robustness to RRL.
Source: Lirias
Authors: Kurt Driessens, Jan Ramon
Institution: KU Leuven
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2148411640