Applying Neural Networks and Genetic Programming to the Game Lost Cities
Abstract
The board game Lost Cities is a non-deterministic game of imperfect information. This makes it difficult to construct a reliable AI player without falling back to computationally expensive Monte Carlo search. We investigated neural networks and genetic programming as part of an alternative approach for constructing an AI player capable of independently developing strategies from recorded games simulated by Monte Carlo search.
Subject
Neural networks
Genetic programming
Posters
Permanent Link
http://digital.library.wisc.edu/1793/79080Description
Color poster with text, charts, and graphs.