The Bump Hunting Using the Decision Tree Combined with the Genetic Algorithm: Extreme-value Statistics Aspect

In difficult classification problems of the z-dimensional points into two groups giving 0-1 responses due to the messy data structure, it is more favorable to search for the denser regions for the response 1 points than to find the boundaries to separate the two groups; this is called the bump hunting. In a series of previous studies, we have shown that a bump hunting method using the decision tree combined with the genetic algorithm is useful for certain customer database, where we have developed a tradeoff curve between the pureness rate and the capture rate. This paper deals with the behavior of the tradeoff curve from a statistical viewpoint.