Research On Rough Set Attribute Reduction Based On Genetic Algorithm

The development of the Database technology and Internet make it possible to collect and store massive quantities of data. When facing to the greatly expanding data, people need more powerful and efficient information processing techniques, which can be used to discover interesting knowledge from massive information and consequently to make decision. So, data mining technology is an effective approach to resolve the problem of abundance in data but scarcity in information. It currently is the research frontier within the information science field .The related researches and applications have greatly improved the ability of decision supporting and it has been deemed to a field that has broad prospect of application in the research of database. Rough set theory is a tool to deal with vague and uncertain data, therefore, it becomes an important frame in data mining. Reduction of knowledge is one of the core contents in rough set theory. Reducted data are more valuable and capable of obtaining accurate knowledge.This thesis studies on attributes reduction algorithm in rough set theory based on genetic algorithm. The main contribution of this thesis includes:Firstly, the thesis reviews the theories and methods of rough set and genetic algorithm systematically, summarizes the classic algorithms of attribute reduction based on rough sets and analyzes the algorithms of attribute reduction based on dicernbility matrix, attribute signficance, dependability and GA.Secondly, as the definition of attribution reduction based on primary discernibility matrix is not the same as the definition of attribution reduction based on positive region, a new definition of discernibility is introduced in this paper; to reduce the searching space of the fitness function and improve the computing efficiency, simple discernibility matrix is improved and the corresponding definition of attribution reduction so that computing complexity is cut down ; by improving the simple genetic algorithm, an attribute reduction algorithm in rough set theory base on immune genetic algorithm is proposed and verified by two database.Finally, base on research of genetic algorithm and rough sets, analysis of the existing primary algorithm of attribute reduction based on the traditonal genetic algorithm, an improved algorithm of rough set attribute reduction base on dependability and genetic algorithm is presented in this paper. At the end of this paper, this new attribute reduction algorithm is applied to forecast maritime accidences and good result is shown through the application.