论文信息 - Multi-type, Self-adaptive Genetic Programming as an Agent Creation Tool

Multi-type, Self-adaptive Genetic Programming as an Agent Creation Tool

This paper describes the application of multitype, self-adaptive genetic programming techniques, implemented in the PushGP and Pushpop systems, to the automatic programming of multi-agent systems. It includes a brief case study of the application of PushGP to a transport network control problem and a demonstration of self-adaptive modularization in a dynamic environment that was developed by Van Belle and Ackley.

Lee Spector | L. Spector

[1] William E. Hart,et al. A Convergence Analysis of Unconstrained and Bound Constrained Evolutionary Pattern Search , 2001, Evolutionary Computation.

[2] C. Titus Brown,et al. Evolutionary Learning in the 2D Artificial Life System "Avida" , 1994, adap-org/9405003.

[3] A. N. Pargellis,et al. The spontaneous generation of digital “Life” , 1996 .

[4] Wolfgang Banzhaf,et al. Genetic Programming: An Introduction , 1997 .

[5] David J. Montana,et al. Strongly Typed Genetic Programming , 1995, Evolutionary Computation.

[6] Peter J. Angeline,et al. Two self-adaptive crossover operators for genetic programming , 1996 .

[7] Thomas S. Ray,et al. Is It Alive or Is It GA? , 1991, ICGA.

[8] Lee Spector,et al. Autoconstructive Evolution: Push, PushGP, and Pushpop , 2001 .

[9] David J. Montana,et al. Evolving control laws for a network of traffic signals , 1996 .

[10] Christopher R. Stephens,et al. Self-Adaptation in Evolving Systems , 1997, Artificial Life.

[11] Terry Van Belle,et al. Code Factoring And The Evolution Of Evolvability , 2002, GECCO.

[12] Lee Spector,et al. Genetic Programming and Autoconstructive Evolution with the Push Programming Language , 2002, Genetic Programming and Evolvable Machines.