Solving identification problem for asynchronous finite state machines using genetic algorithms

A Genetic Algorithm, embedded in a simulation-based method, is applied to the identification of Asynchronous Finite State Machines. Two different coding schemes and their associated crossover operations are examined. It is shown that one operator / coding pair outperforms the other in that the scheme reduces noticeably the production of invalid chromosomes thus increasing the efficiency and the convergence rate of the evolution process. Categories and Subject Descriptors I.6.5 [Model Development]: Modeling methodologies. General Terms Algorithms, Performance, Experimentation. Keywords Model Identification, Asynchronous Finite State Machines, Genetic Algorithms, Coding Schemes.

• Type: Conference
• Year: 2006
• Where: GECCO
• Authors: Xiaojun Geng
• Keywords: Asynchronous Finite State, Coding Schemes, Finite State Machines, GECCO 2006, Optimization