Data Mining Approach for Decision and Classification Systems Using Logic Synthesis Algorithms

Abstract

This chapter discusses analogies between decision system and logic circuit. For example, the problem of data redundancy in decision system is solved by minimizing the number of attributes and removing redundant decision rules which is analogous to the argument reduction method for logic circuits. Another issue associated with the field of data mining lies in the induction of decision rules which in result provide a basis for decision-making tasks. A similar algorithm in logic synthesis is called minimization of Boolean function. An issue of reduction of the capacity required to memorize a decision table is solved by disassembling this table to the subsystems in such a way that the original one can be recreated through hierarchical decision making. In logic synthesis it is called functional decomposition and is used for efficient technology mapping of logic circuits. Due to different interpretation and application these tasks seem totally different, however the analogies allow logic synthesis algorithms to be used in the field of data mining. Moreover, by applying specialized logic synthesis methods, these three issues, i.e. feature reduction, rule induction, and hierarchical decision making, can be successfully improved.