智慧城市网格管理事件模式挖掘与预测
Urban Grid Management Incidents Pattern Mining and Prediction
吴 俊
上海信投建设有限公司 副总经理，高级工程师
金耀辉
上海交通大学人工智能研究院 教授，博士生导师，博士
王杰艺
上海交通大学电子信息与电气工程学院 硕士研究生
摘要: 城市化进程的快速发展给城市管理带来了巨大的挑战，城市问题事件的管理和预警已成为城市可持续发展的一个重要组
成部分。多变量时间序列分析预测模型RBTA可以挖掘城市管理事件发展的基本趋势、周期性、异常事件以及不同城市管
理事件之间的耦合关系。使用来自于上海市徐汇区网格中心的真实城市管理事件数据及对模型进行评估，该模型的平均
拟合均方根误差为0.12，平均预测均方根误差为0.15，预测准确率比现有方法中最好的方法高4.9%。
Abstract: The rapid development of urbanization has brought great challenges to the management of cities. The management and early warning
of urban incidents have become an important part of urban sustainable development. This paper proposes RBTA, a multivariate timeseries
model, to find the patterns including basic trend, seasonality, irregular components and relationship among different incidents.
We evaluate our model on the real dataset from the downtown area of Shanghai, one of the biggest metropolitan of the world.
The average forecasting root mean squared error (RMSE) is 0.15, which decreases 4.9% comparing to the best one of the existing
methods.
关键词：智慧城市、时间序列、 预测
Keyword: Smart city ，Time-series，Forecast
中图分类号：TU981
文献标识码: A
[1]HAMBLEN M. Just what is a smart city?[EB/
OL]. (2015-10-01) [2018-03-06]. https://www.
computerworld.com/article/2986403/internet-ofthings/
just-what-is-a-smart-city.html.
[2]GIFFINGER R, FERTNER C, KRAMAR H, et al.
City-ranking of European medium-sized cities[R].
2014.
[3]DEAKIN M. Smart cities: governing, modeling,
and analyzing the transition[M]. Oxford: Routledge,
2013.
[4]BARBER B R. If mayors ruled the world:
dysfunctional nations, rising cities[M]. New Haven:
Yale University Press, 2013: 186.
[5]PIERRE J. Models of urban governance: the
institutional dimension of urban politics[J]. Urban
Affairs Review, 1999, 34(3): 372-396.
[6]NAM T, PARDO T A. Smart city as urban innovation:
focusing on management, policy, and context[C]//
Proceedings of the International Conference on
Theory and Practice of Electronic Governance.
Tallinn, Estonia, September, 2011: 185-194.
[7]GIL-GARCIA J R. Enacting electronic government
success: an integrative study of governmentwide
websites, organizational capabilities, and
institutions[J]. Integrated, 2012, 22(21): 3144-3150.
[8]KORTUEM G, KAWSAR F, SUNDRAMOORTHY
V, et al. Smart objects as building blocks for the
Internet of things[J]. IEEE Internet Computing, 2009,
14(1): 44-51.
[9]MATSUBARA Y, SAKURAI Y, FALOUTSOS
C. AutoPlait: automatic mining of co-evolving
time sequences[C]//ACM SIGMOD International
Conference on Management of Data. 2014: 193-204.
[10]LI L, PRAKASH B A, FALOUTSOS C.
Parsimonious linear fingerprinting for time series[J].
Proceedings of the Vldb Endowment, 2010, 3(1-2):
385-396.
[11]MATSUBARA Y, SAKURAI Y, FALOUTSOS C,
et al. Fast mining and forecasting of complex timestamped
events[C]//ACM SIGKDD International
Conference on Knowledge Discovery and Data
Mining. 2012: 271-279.
[12]MATSUBARA Y, SAKURAI Y, PANHUIS W G
V, et al. FUNNEL: automatic mining of spatially
coevolving epidemics[C]//Proceedings of the 20th
ACM SIGKDD Conference, New York City, August
24-27, 2014: 105-114.
[13]LIVERA A M D, HYNDMAN R J, SNYDER R
D. Forecasting time series with complex seasonal
patterns using exponential smoothing[J]. Monash
Econometrics & Business Statistics Working Papers,
2011, 106(496): 1513-1527.
[14]MATSUBARA Y, SAKURAI Y, FALOUTSOS C.
The web as a jungle: non-linear dynamical systems
for co-evolving online activities[J]. 2015, 24(3): 721-
731.
