The general reduction algorithm of information system on heterogeneous data

For the real information systems and decision-making system, data types are often heterogeneous data types problem, the paper gives the corresponding data structures, based on neighborhood rough model and the particle swarm optimization ideas, put forward the general reduction algorithm. By the classical data sets and three UCI hybrid data set reduction, the results show that the algorithm´s effectiveness and feasibility. In addition, the experiments also show that the algorithm can solve some problems that the existing heuristic algorithms can not solve.