A Sequential Inverse Classification Problem
This paper proposes a sequentialinverse classification problem (SICP) where a series of improvements are sought over time to classify a case from the currently classified undesirable class into another desirable class.These problems occur at personal and at organizational levels. Many insolvency, reengineering,and continuous quality improvement problems may be formulated as inverse sequential classification problems. Both linear and non-linear data envelopment analysis based procedures can be used to solve these SICP problems. This paper proposes and illustrates application of data envelopment analysis to solve SICP problems.
Index Terms - Inverse classification, Data Mining, Analytical Geometry, Linear Programming.