Exploiting HW Acceleration for Classifying Complex Test Program Generation Problems
of Evolutionary Computing: EvoWorkshops 2004 proceedings, Coimbra (Portugal), April 5-7 2004, pp. 230-239
KEYWORDS:
Approximate Methods,
Evolutionary Algorithms,
Genetic Programming,
Micro-Processors,
MicroGP
ABSTRACT
This paper presents a complete framework to examine, evaluate and characterize an evolutionary test-program generation problem. Our methodology is based on a local analysis of the relationships between genotype and fitness. Furthermore we propose the adoption of a hardware accelerator device for speeding up program cultivation. A commercial, complex microprocessor was used as case study. Experimental analysis allows discovering the characteristics of the specific task and foreseeing the behavior of the test-program generation.
[SSVi04] E. Sanchez, G. Squillero, M. Violante, "Exploiting HW Acceleration for Classifying Complex Test Program Generation Problems," of Evolutionary Computing: EvoWorkshops 2004 proceedings, Coimbra (Portugal), April 5-7 2004, pp. 230-239 |
|
