Cost-sensitive techniques for fuzzy rule-based pattern classification

This paper surveys cost-sensitive fuzzy rule-based systems for pattern classification. Weighted training patterns are used to construct cost-sensitive fuzzy rule-based systems. A fuzzy classification system is constructed from a given set of training patterns. It is assumed that a weight is assigned to each training pattern a priori. The weight of training patterns can be determined based on their distribution. We show an example of weight assignment for a two-dimensional problem.