The Genetic Algorithm can be modified in different ways. Even Goldberg's book, Genetic Algorithms in Search, Optimization and Machine Learning, specifies various enhancements that could be done for genetic search.
Which is the exact method implemented in Weka for genetic search for attribute selection? The software does refer Goldberg's book, but I'm still not sure about its deeper specifics.
Some of which are
- Precise fitness evaluation function
- If fitness modification to add weight to diversity is implemented
- Structure encoding of the chromosome
- Type/variant of crossover implemented: single point, multi-point, and which point to be exact.
In feature selection, Genetic Algorithm (GA) is used as a random selection algorithm, Capable of effectively exploring large search spaces, which is usually required in case of attribute selection. For instance; if the original feature set contains N number of features, the total number of competing candidate subsets to be generated is 2N, which is a huge number even for medium-sized N. Further, unlike many search
algorithms, which perform a local, greedy search, GAs performs a global search
I agree with Dr. Aguilar-Ruiz, but only as last resort. I presume that some may have already explored code and might provide a solution.
The GA in WEKA is from Goldberg (1989)
Article Genetic Algorithms In Search, Optimization, and Machine Learning
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology