Evolution Programs and Heuristics

1996 ◽  
pp. 307-327 ◽  
Author(s):  
Zbigniew Michalewicz
Keyword(s):  
Evolution Programs

scholarly journals Pioneer: A New Tool for Coding of Multi-Level Finite State Machines Based on Evolution Programming

VLSI Design ◽  
1994 ◽  
Vol 2 (2) ◽  
pp. 105-116
Author(s):  
S. Muddappa ◽  
R. Z. Makki ◽  
Z. Michalewicz ◽  
S. Isukapalli
Keyword(s):  
Finite State Machines ◽  
Mutation Rates ◽  
Natural Phenomenon ◽  
State Machines ◽  
State Assignment ◽  
Wide Range ◽  
Finite State ◽  
Multi Level ◽  
Crossover And Mutation ◽  
Evolution Programs

In this paper we present a new tool for the encoding of multi-level finite state machines based on the concept of evolution programming. Evolution programs are stochastic adaptive algorithms, based on the paradigm of genetic algorithms whose search methods model some natural phenomenon: genetic inheritance and Darwinian strife for survival. Crossover and mutation rates were tailored to the state assignment problem experimentally. We present results over a wide range of MCNC benchmarks which demonstrate the effectiveness of the new tool. The results show that evolution programs can be effectively applied to state assignment.

An Object Oriented Library for Evolution Programs

2004 ◽  
pp. 351-360
Author(s):  
Jarosław Żola ◽  
Łukasz Łaciński ◽  
Roman Wyrzykowski
Keyword(s):  
Object Oriented ◽  
Evolution Programs

scholarly journals A Hierarchy of Evolution Programs: An Experimental Study

1993 ◽  
Vol 1 (1) ◽  
pp. 51-76 ◽  
Author(s):  
Zbigniew Michalewicz
Keyword(s):  
Experimental Study ◽  
General Purpose ◽  
Specific Knowledge ◽  
Trade Offs ◽  
Evolution Programs ◽  
Effective Representation ◽  
Better Than

In this paper we present the concept of evolution programs and discuss a hierarchy of such programs for a particular problem. We argue that (for a particular problem) stronger evolution programs (in terms of the problem-specific knowledge incorporated in the system) should perform better than weaker ones. This hypothesis is based on a number of experiments and a simple intuition that problem-specific knowledge enhances an algorithm's performance; at the same time it narrows the applicability of an algorithm. Trade-offs between the effort of finding an effective representation for general-purpose evolution programs and the effort of developing more specialized systems are also discussed.

Sign in / Sign up

Export Citation Format

Share Document