- author = "Michael O'Neill and Conor Ryan",
- title = "Automatic Generation of Caching Algorithms",
- booktitle = "Evolutionary Algorithms in Engineering and Computer Science",
- year = "1999",
- editor = "Kaisa Miettinen and Marko M. M{\"{a}}kel{\"{a}} and Pekka Neittaanm{\"{a}}ki and Jacques Periaux",
- pages = "127--134",
- address = "Jyv{\"{a}}skyl{\"{a}}, Finland",
- publisher_address = "Chichester, UK",
- month = "30 " # may # " - 3 " # jun,
- publisher = "John Wiley \& Sons",
- keywords = "genetic algorithms, genetic programming, grammatical evolution",
- isbn13 = "978-0-471-99902-7",
-
URL = "
http://www.mit.jyu.fi/eurogen99/papers/oneill.ps",
- size = "8 pages",
-
notes = "EUROGEN'99
http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0471999024.html
Claims performance better than human coded 'Least Recently Used' LRU algorithm. Provides indexed memory info[] but evolved solutions don't use it. 'as closely as possible to the procedures outlined in \cite{Paterson:1997:ecacGP}' Small cachesize (20), large CACHESIZE 200 46percent better than LRU. GE gives less over fitting. Test data like \cite{Paterson:1997:ecacGP}",
