The Bump Hunting Method and Its Accuracy Using the Genetic Algorithm with Application to Real Customer Data

Suppose that we are interested in searching for denser regions showing response 1 with many feature variables (explanation variables) in a z-dimensional space, where each point is assigned response 1 or response 0 as its target value; such a region is called the bump. In a series of previous studies, we have shown that the bump hunting method using the decision tree combined with the genetic algorithm is useful for certain smaller simulated data case mimicked to a real customer case. We have developed a trade-off curve with its accuracy evaluation between the pureness rate and the capture rate to the simulated data. This paper deals with a real customer data case, and we have found that it is crucial to know the relation between the number of feature variables and the number of samples.