Incremental rules mining for information compression matrix algorithm

Dynamic rules acquisition is a topic of general interest in the field of knowledge discovery. Existing matrix algorithm cannot satisfy completely, rules acquisition and quick updating for changing information. It is necessary to be extended in the face of this challenge. Rough set theory (RST) is a new efficient tool for rules acquisition. An innovative incremental mining RST-based algorithm of information compression matrix (ICM) is proposed. Relative core and relative reduction are defined, and incremental algorithm of rules acquisition based on ICM is presented. Rules acquisition on the basis of existing rules is to update rules and rules´ parameters dynamically. It avoids traversing all attributes and records repeatedly, and reduces the time and space complexity of algorithm. Experimental results verify the efficiency and validity of the algorithm. The proposed method serves the operating optimization of ethylene cracking furnace quite well in the process industry.