schrum2 / MM-NEATv2

MM-NEAT version 2.0 is no longer supported. Please get MM-NEAT 3+ from https://github.com/schrum2/MM-NEAT
Other
11 stars 5 forks source link

This version of MM-NEAT is not longer being developed. Please get the latest version from https://github.com/schrum2/MM-NEAT

MM-NEAT version 2.0

Copyright (c) 2014 The University of Texas at Austin and 2016 Southwestern University. All rights reserved. Refer to LICENSE.txt for detailed license information. Also see copyright.txt for copyright information about the included Ms. Pac-Man code.

WEBPAGE

http://nn.cs.utexas.edu/?mm-neat Although, the latest version is on GitHub: https://github.com/schrum2/MM-NEATv2

ABOUT

MM-NEAT stands for Multiobjective Modular Neuro-Evolution of Augmenting Topologies. It is inspired by the original NEAT, but also incorporates multiobjective evolution via NSGA-II, and supports several forms of modular neural networks. Support for the fitness shaping technique Targeting Unachieved Goals (TUG) is also included. Indirect encoding via HyperNEAT is also supported, as is interactive evolution of several types of interesting artifacts (pictures, sounds, animations). The code was originally developed by Jacob Schrum (schrum2@southwestern.edu) while at the University of Texas at Austin, but has since been improved upon by several undergraduate students at Southwestern University in Georgetown, TX, where Dr. Schrum is currently a professor in the department of Math and Computer Science. Links to publications and demos further explaining the code are available at the official webpage:

http://nn.cs.utexas.edu/?mm-neat Also on Dr. Schrum's personal webpage: http://www.southwestern.edu/~schrum2/

More information on NEAT is available in:

K. O. Stanley and R. Miikkulainen, "Evolving Neural Networks Through Augmenting Topologies." Evolutionary Computation, 10(2):99-127, 2002.

Information on NSGA-II is available in:

K. Deb, S. Agrawal, A. Pratab, and T. Meyarivan, "A fast elitist nondominated sorting genetic algorithm for multi-objective optimization: NSGA-II". PPSN VI, pp. 849-858, 2000.

A precursor to MM-NEAT is the BREVE Monsters software package, available at:

http://nn.cs.utexas.edu/?brevemonsters

This code was developed primarily to evolve multimodal behavior in Ms. Pac-Man, and therefore includes (modified) code for the Ms. Pac-Man simulator created for the Ms. Pac-Man vs. Ghosts Competitions. The original version of this code does not seem to be available anymore, but a newer version associated with the latest competition can be downloaded at:

http://www.pacmanvghosts.co.uk/

For further instructions on how to run this code, see TUTORIAL.txt.

For information on the different types of interactive evolution tasks in the code, see INTERACTIVE_EVOLUTION.txt

FOR MORE INFORMATION CONTACT

schrum2@southwestern.edu