Research Of Attribute Reduction Method Based On Heuristic Genetic Algorithm

With the development of technology of acquiring the spatial data, it becomes more and more important to find knowledge from a large number of spatial data automatically, quickly, and efficiently. The concept, characteristics of spatial data mining and the knowledge types which can be found by spatial data mining are introduced by this paper. And the system structure and basic course of spatial data mining are shown. Then, the method of the attribute reduction in the decision table which is a part of the data pretreatment in spatial data mining is discussed as emphasis.According to the characteristic of rough set theory, genetic algorithm as well as decision table attribute reduction, through analysis of the primary algorithm of attribute reduction which based on the traditional genetic algorithm in rough set, a new algorithm of attribute reduction which is based on heuristic genetic algorithm in rough set (HGAAR algorithm) is put forward in this paper. There are two important parts which are improved in the algorithm. The first one is that the new algorithm uses attribute core in decision table as a restriction to improve the binary code initial population which is produced stochastically in the traditional genetic algorithm. This improvement can strengthen the local search ability of genetic algorithm, shorten the calculation time of the algorithm, and raise the accuracy of the results of the attribute reduction in decision tables. The other one based with the genetic operators (which includes selection operator, crossover operator and mutation operator), the new algorithm increases a correction operator. In this operator, the algorithm uses attribute significance as heuristic information which can guide the algorithm carrying out in the space of feasible results, and the attribute which can make more important influence is prevented to be lost. And, more efficient attributes can be provided for the spatial data mining. Finally, an experiment is shown to prove that the HGAAR algorithm is more excellence than...