@inbook{200f02fee2bd4714a593355fb110e817,
title = "Evolving parameterised policies for stochastic constraint programming",
abstract = "Stochastic Constraint Programming is an extension of Constraint Programming for modelling and solving combinatorial problems involving uncertainty. A solution to such a problem is a policy tree that specifies decision variable assignments in each scenario. Several solution methods have been proposed but none seems practical for large multi-stage problems. We propose an incomplete approach: specifying a policy tree indirectly by a parameterised function, whose parameter values are found by evolutionary search. On some problems this method is orders of magnitude faster than a state-of-the-art scenario-based approach, and it also provides a very compact representation of policy trees.",
author = "Steven Prestwich and Tarim, \{S. Armagan\} and Roberto Rossi and Brahim Hnich",
year = "2009",
doi = "10.1007/978-3-642-04244-7\_53",
language = "English",
isbn = "3642042430",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "684--691",
booktitle = "Principles and Practice of Constraint Programming - CP 2009 - 15th International Conference, CP 2009, Proceedings",
note = "15th International Conference on Principles and Practice of Constraint Programming, CP 2009 ; Conference date: 20-09-2009 Through 24-09-2009",
}