Abstract: Data mining techniques can be used to find out potential and useful knowledge from the vast amount of data,and it plays a new significant role to the stored data in the info-times.With the rapid development of the data mining techniques,the technique of grid clustering,as important parts of data mining,are widely applied to the fields such as pattern recognition,data analysis,image processing,and market research.Research on grid clustering algorithms has become a highly active topic in the data mining research.In this thesis,the author presented the theory of data mining,and deeply analyzes the algorithms of grid clustering.Based on the analysis of traditional grid clustering algorithms,we advanced some improved grid clustering algorithms that can enhance the quality and efficiency of grid clustering compared with the traditional grid clustering algorithms.Based on the analysis of traditional algorithms for multi-density,we advanced a grid-based clustering algorithm for multi-density(GDD).The GDD is a kind of the multi-stage clustering that integrates grid-based clustering,the technique of density threshold descending and border points extraction.As shown in the research,GDD algorithm can not only clusters correctly but find outliers in the dataset,and it effectively solves the problem that traditional grid algorithms can cluster only or find outliers only.The precision of GDD algorithm is better than that of SNN.The GDD algorithm works well for even density dataset and lots of multi-density datasets；it can discover clusters of arbitrary shapes；it isn’t sensitive to the input order of noises and outliers data,but it is imperfect to cluster on some multi-density datasets.

Key words: Grid clustering,Density threshold descending,Multi-stage clustering