基于 T(X )参与度的负co-location模式挖掘算法

Negative co-location pattern mining algorithm based on T(X)

空间co-location模式是一组在空间中频繁并置的空间特征的子集.负co-location模式从非频繁的空间co-location模式中产生.一般来说很难计算和挖掘频繁的负co-location模式.频繁负co-location模式中有较强的应用价值,如发现外来物种入侵,自然界植被生长规律等.现有对负co-location模式研究不全面且挖掘算法的数量屈指可数.针对该问题,提出了T(X)下的负co-location模式的参与度度量方法,并分析了此度量的合理性、可行性和简便性;其次,利用此度量,可以发现负模式中隐含的“团爆炸”现象,而之前的度量方式不能发现此现象.提出了基于T(X)参与度度量的负co-location模式挖掘算法.最后,实验结果表明,在其他条件不变的情况下,该算法可以挖掘数量更少且更具负相关性的频繁负co-location模式.

A spatial co-location pattern is a subset of a set of spatial features that are frequently collocated in space.The negative co-location pattern arises from the infrequent spatial co-location pattern.It is often difficult to calculate and mine frequent negative co-location patterns.The frequent negative co-location model has strong application value,such as the discovery of alien species invasion,natural vegetation growth law,etc.The existing research on negative co-location pattern is incomplete and the number of mining algorithms is very small.Aiming at this problem,firstly,TNPI is proposed,and the rationality,feasibility and simplicity of this measurement are analyzed.Secondly,the phenomenon of“mass explosion”implied in the negative pattern can be found using this metric,which cannot be found in previous measurements.A negative co-location pattern mining algorithm based on T(X)engagement metric is proposed.Finally,experimental results show that the algorithm can mine fewer and more negatively correlated frequent negative co-location patterns without other conditions being constant.