Affiliations: Lisboa, Portugal / Funchal, Portugal
Note: [1] Faculdade de Ciências da Universidade de Lisboa (FCLTL), Departamento de Informática, Campo Grande, 1749-016 Lisboa, Portugal, email: tl@di.fc.ul.pt
Note: [2] Universidade da Madeira, Departamento de Matematica e Engenhiaria (INEGC-ID), 9000-390 Funchal, Portugal, email: pcampos@uma.pt
Abstract: The article describes ABALEARN, a self-learning Abalone program capable of automatically reaching an intermediate level of play. This is achieved without expert-labelled training examples, deep searches, and exposure to competent play. Our approach is based on a reinforcement-learning algorithm that is risk seeking, since defensive players in Abalone tend to postpone a game endlessly. We show that risk sensitivity allows a successful self-play training. Moreover, we propose a set of features that seem relevant for achieving a rather skilled level of play. We evaluate our approach using a fixed heuristic opponent as a benchmark. We pit our agents against human players on-line and compare samples of our agents at different times of training.
DOI: 10.3233/ICG-2003-26402
Journal: ICGA Journal, vol. 26, no. 4, pp. 219-228, 2003
