發(fā)布時間：2019-01-06 12:03

【摘要】：探討了位置數(shù)據(jù)異常分析的一種新思路,給出了面向異常的時空表達模型,建立了基于區(qū)域位置點統(tǒng)計的異常檢測和分析框架,探尋異常在事件影響下的時空發(fā)展趨勢和分布特征。通過紐約兩個月的出租車出行數(shù)據(jù)的實驗分析驗證了框架的有效性,并發(fā)現(xiàn)一些有趣的規(guī)律,為公共安全、交通管控應(yīng)急、城市功能區(qū)分析等提供指導(dǎo)信息。
[Abstract]:In this paper, a new idea of abnormal analysis of position data is discussed, and an anomaly oriented spatio-temporal representation model is presented, and an anomaly detection and analysis framework based on the statistics of regional location points is established. To explore the temporal and spatial trends and distribution characteristics of anomalies under the influence of events. The validity of the framework is verified by the experimental analysis of two months' taxi travel data in New York, and some interesting rules are found to provide guidance information for public safety, traffic control and emergency response, analysis of urban functional areas, etc.
【作者單位】： 信息工程大學(xué);61206部隊;
【基金】：國家自然科學(xué)基金項目(41571394)
【分類號】：P208
，

本文編號：2402763