- author = "Norman Paterson and Mike Livesey",
- title = "Evolving caching algorithms in C by genetic programming",
- booktitle = "Genetic Programming 1997: Proceedings of the Second Annual Conference",
- editor = "John R. Koza and Kalyanmoy Deb and Marco Dorigo and David B. Fogel and Max Garzon and Hitoshi Iba and Rick L. Riolo",
- year = "1997",
- month = "13-16 " # jul,
- keywords = "genetic algorithms, genetic programming, GADS, BNF grammar, GAGS-1.0, linear GP",
- pages = "262--267",
- address = "Stanford University, CA, USA",
- publisher_address = "San Francisco, CA, USA",
- publisher = "Morgan Kaufmann",
-
URL = "
http://www.dcs.st-and.ac.uk/~norman/Pubs/cache.ps.gz",
- size = "6 pages",
- abstract = "This paper outlines current work in ontogenic-mapped and language abstracted genetic programming. Developments in the genetic algorithm for deriving software (GADS) technique are described, and tested in a series of experiments to generate caching algorithms in C. GADS quickly finds over-fitted solutions which perform better than designed solutions but only in one niche. The need for a scalable approach in GADS to deal with language definitions involving more productions is demonstrated.",
-
notes = "GP-97. Test data also used in
\cite{oneill:1999:AGCA}
Production weights on BNF grammar. No mutation. GP returns long victim which forced into legal cache line range by wrapper which gives cache line of data to be ejected from cache. Evolved C code decides who to evict from memory access cache. INFO array, read() write_x, small_x, large_x, random_x, counter, CACHESIZE, div, rem. phenotype.cc Linear bitstring (2500 or 3000 bits) GA (5 or 6 bits per BNF production. Ontogenic mapping. Repair. BNF text converted to C++ data structure. Number of productions varies with each BNF rule (up to 2**6). Uses multiple identical replicated productions of some BNF rules. Each C++ phenotype compiled for fitness calculation using trace file (Flanagan, 1992).",
