Title :
Solving randomly generated fuzzy constraint networks using evolutionary/systematic hill-climbing
Author :
Bowen, James ; Dozier, Gerry
Author_Institution :
Dept. of Comput. Sci., Nat. Univ. of Ireland, Cork, Ireland
Abstract :
This paper introduces an evolutionary/systematic hybrid which combines the concept of evolutionary hill-climbing search with the systematic search concept of arc revision to form a hybrid that quickly find solutions to fuzzy constraint satisfaction problems. This hybrid outperforms a modified version of a well known hill-climber, the iterative descent method, on a test suite of 500 randomly generated fuzzy constraint networks
Keywords :
constraint handling; fuzzy set theory; genetic algorithms; search problems; arc revision; evolutionary algorithm; evolutionary/systematic hybrid; fuzzy constraint networks; fuzzy constraint satisfaction problems; fuzzy set theory; hill-climbing search; systematic search; Computer science; Control engineering; Evolutionary computation; Fuzzy control; Fuzzy sets; Fuzzy systems; Machine learning; NASA; Robustness; Testing;
Conference_Titel :
Fuzzy Systems, 1996., Proceedings of the Fifth IEEE International Conference on
Conference_Location :
New Orleans, LA
Print_ISBN :
0-7803-3645-3
DOI :
10.1109/FUZZY.1996.551746
