查看文章

Contrast set mining is one of the most important tasks in the supervised descriptive pattern mining field. It aims at finding patterns whose frequencies differ significantly among sets of data under contrast. This chapter introduces therefore the contrast set mining problem as well as similarities and differences with regard to related techniques. Then, the problem is formally described and the most widely used metrics in this task are mathematically defined. In this chapter, the most important approaches in this field are analysed, including STUCCO and CIGAR among others. Finally, some additional proposals in the field are also described.