- author = "K. C. Tan and A. Tay and T. H. Lee and C. M. Heng",
- title = "Mining multiple comprehensible classification rules using genetic programming",
- booktitle = "Proceedings of the 2002 Congress on Evolutionary Computation CEC2002",
- editor = "David B. Fogel and Mohamed A. El-Sharkawi and Xin Yao and Garry Greenwood and Hitoshi Iba and Paul Marrow and Mark Shackleton",
- pages = "1302--1307",
- year = "2002",
- publisher = "IEEE Press",
- publisher_address = "445 Hoes Lane, P.O. Box 1331, Piscataway, NJ 08855-1331, USA",
- organisation = "IEEE Neural Network Council (NNC), Institution of Electrical Engineers (IEE), Evolutionary Programming Society (EPS)",
- ISBN = "0-7803-7278-6",
- month = "12-17 " # may,
- notes = "CEC 2002 - A joint meeting of the IEEE, the Evolutionary Programming Society, and the IEE. Held in connection with the World Congress on Computational Intelligence (WCCI 2002)",
- keywords = "genetic algorithms, genetic programming, IF-THEN rule evolution, application domains, artificial immune system-like memory vector, benchmark data sets, concept mapping technique, covering algorithm, data mining, fitness evaluation, multiple comprehensible classification rules, redundant rule removal, simulation, tree representation, data mining, pattern classification, programming, redundancy",
-
DOI = "
doi:10.1109/CEC.2002.1004431",
- abstract = "Genetic Programming (GP) has been emerged as a promising approach to deal with classification task in data mining. This work extends the tree representation of GP to evolve multiple comprehensible IF-THEN classification rules. In the paper, we introduce a concept mapping technique for fitness evaluation of individuals. A covering algorithm that employs an artificial immune system-like memory vector is used to produce multiple rules as well as to remove redundant rules. The proposed GP classifier is validated upon nine benchmark datasets and the simulation results confirm the viability and effectiveness of the GP approach for solving data mining problems in a wide spectrum of application domains.",
-
notes = "Michigan approach.
GPc, groovy Java GP (gjprog), WEKA. problem specific population sizes 10-100 and w_1 and w_2.
",
