一种基于KD－Tree的同位模式挖掘算法研究

Co-location Pattern Mining Algorithm Study Based on KD-Tree

空间同位模式挖掘是空间数据挖掘中一种重要的方法。空间同位模式是指在地理空间中，由于空间接近性而频繁的聚集在一起的空间特征的集合。提出了一种基于KD树的同位模式挖掘算法，算法可以分为两个基本步骤：邻域的确认和同位模式的剪枝过程。最后通过实验表明了算法的准确性与全面性。

Co-location patterns mining is an important spatial data mining method.Co-location patterns is defined geographic space,because of the space close to nature and frequent together as a collection of spatial features.KD-tree based on the same bit pattern mining algorithm,the algorithm can be divided into two basic steps： Recognized neighborhood and pruning process of co-location pattern.The experiments show the algorithm accuracy and comprehensiveness.