- author = "Kim Harries and Peter Smith",
- title = "Exploring Alternative Operators and Search Strategies in 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",
- pages = "147--155",
- address = "Stanford University, CA, USA",
- publisher_address = "San Francisco, CA, USA",
- publisher = "Morgan Kaufmann",
-
URL = "
http://www.cs.ucl.ac.uk/staff/W.Langdon/ftp/papers/harries.gp97_paper.ps.gz",
-
notes = "GP-97
even-4-parity, even-5-parity, artificial ant (Santa Fe trail), regression of x^4-3x^3+9x^2-27x Depth-based crossover (depth fair, SameDepths and DiffDepths) NoBias and combinations of crossovers. SameDepths does badly on even-5-parity otherwise crossovers similar to each other. 58,100 runs
Several different types of mutation (and combination of mutation) used as stochastic {"}hill climbers{"}.
Steady state, tournament size=2, limit of 1000 nodes, kinnear's Hoist and mutation both at 1 percent.
GP mutation generally good hill climber, small and self-crossover generally awful. See sec 4 discussion. )",
