论文部分内容阅读
采用Apriori算法将大连市出租车的时空数据进行聚类分析与关联分析,对城市交通流进行预测。阐述了出租车时空数据挖掘的基本过程与思想方法,着重介绍了对于时空数据在此挖掘方法下的特征架构结构与分析结果。通过对车载GPS采集到的原始数据进行阈值界限预处理,确定其上下阈值界限,并对处理后的交通流数据进行挖掘预测,将改进后Apriori算法应用于时空数据的挖掘中,对区域内的不同时段的路段分别进行挖掘处理,生成满足约束条件的频繁集,并找寻出其中的强关联性,使时空大数据的挖掘效果更加合理,预测结果更加准确。通过挖掘分析后的数形展示,对于管理道路交通的拥堵将提供有效的决策依据。
The Apriori algorithm is used to cluster and correlate the taxi data in Dalian to predict the urban traffic flow. The basic process and thinking method of taxi spatio-temporal data mining are expounded. The structure and analysis results of the characteristic architecture of spatio-temporal data under this mining method are emphatically introduced. By pretreating the original data collected by GPS on the vehicle to determine the upper and lower threshold limits and mining the predicted traffic flow data, the improved Apriori algorithm is applied to the mining of spatio-temporal data, The road segments at different time periods are respectively subjected to mining processing to generate frequent sets satisfying constraints and to find the strong correlation among them so as to make the mining effect of the large-scale spatiotemporal data more reasonable and the prediction results more accurate. By mining the analyzed numerical representation, congestion control for road traffic will provide an effective basis for decision making.