基于布朗桥模型的时空同现模式分析方法  

Co-Occurrence Pattern Analysis Method Base on the Brownian Bridge Model

作　　者：邓超[1,2] 罗泽[1] 阎保平[1] 

机构地区：[1]中国科学院计算机网络信息中心,北京100190 [2]中国科学院大学,北京100049

出　　处：《科研信息化技术与应用》2014年第3期50-58,共9页E-science Technology & Application

基　　金：国家自然科学基金(中美软件合作研究项目)(61361126011);中国科学院信息化专项(XXH12504-1-06)

摘　　要：同现模式挖掘一直是时空数据挖掘的重要部分,现有的同现模式定义无法准确地描述同现实例。为了解决该问题,本文提出了一种从概率角度来描述同现模式的思路,并提出了基于布朗桥的同现概率分布建模方法。该方法利用布朗桥模型,对移动对象的轨迹分布分别建模,再计算对象与对象同现的概率分布。该方法从概率的角度解释了同现发生的可能性,对同现实例的描述更加精确。最后,本文将该方法应用在青海湖斑头雁迁徙的时空数据上,对同现分布进行建模,发现高概率同现分布的区域和时间,并跟踪了移动对象随时间变化对应的高概率同现区域的变化情况。Co-occurrence pattern mining has always been an important part of spatio-temporal data mining, the existing deifnition of co-occurrence patterns cannot describe co-occurrence cases accurately. To solve this problem, this paper proposed an idea to describe co-occurrence cases in a probabilistic perspective and a method for probability distribution modeling base on Brownian bridge model,which uses the Brownian Bridge to model trajectory of each moving object, and then calculates the probability distribution of co-occurrence between objects. This method explains the probability of a co-occurrence case ,and can describe the co-occurrence more accurately. After all, this paper applies this algorithm on the data of bar-headed goose in the Qinghai Lake Area to model the probability density distribution of co-occurrence patterns, to ifnd co-occurrence regions and time with high probability, and to track the change of high probability co-occurrence regions as time goes on.

关 键 词：时空数据 同现模式 布朗桥模型 概率密度分布 

分 类 号：P208[天文地球—地图制图学与地理信息工程]