Learning First-order Relations from Noisy Databases using Genetic Algorithms

In knowledge discovery from databases, we emphasize the need for learning from huge, incomplete and imperfect data sets (Piatetsky-Shapiro and Frawley, 1991). To handle noise in the problem domain, existing learning systems avoid overfitting the imperfect training examples by excluding insignificant patterns. The problem is that these systems use a limiting attribute-value language for representing the training examples and induced knowledge. Moreover, some important patterns are ignored because they are statistically insignificant. This paper describes a system called GLPS that combines Genetic Algorithms and a variation of FOIL (Quinlan, 1990) to learn first-order concepts from noisy training examples. The performance of GLPS is evaluated on the chess endgame domain. A detail comparison to FOIL is accomplished and the performance of GLPS is significantly better than that of FOIL. This result indicates that the Darwinian principle of natural selection is a plausible noise handling method which can avoid overfitting and identify important patterns at the same time.