Prediction of next destinations from irregular patterns

被引:8
作者
Boukhechba, Mehdi [1 ]
Bouzouane, Abdenour [1 ]
Gaboury, Sebastien [1 ]
Gouin-Vallerand, Charles [2 ]
Giroux, Sylvain [3 ]
Bouchard, Bruno [1 ]
机构
[1] Univ Quebec Chicoutimi UQAC, LIARA Lab, Chicoutimi, PQ, Canada
[2] Univ Quebec TELUQ, Montreal, PQ, Canada
[3] Univ Sherbrooke, Sherbrooke, PQ, Canada
基金
加拿大魁北克医学研究基金会;
关键词
Human activities; Activity prediction; Online association rules; Concept drift;
D O I
10.1007/s12652-017-0519-z
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Few decades ago, understanding human behaviors was considered as a mystery where predicting people's future was impossible. Many changes have been noticed since that era. Thanks to current advances in location tracking technology and data mining techniques, predicting users' behaviors has become possible. In this paper we present a new algorithm to online predict users' next visited locations that not only learns incrementally the users' habits, but also detects and supports the drifts in their patterns. Our original contribution includes a new algorithm for online association rules mining that supports the concept drift.
引用
收藏
页码:1345 / 1357
页数:13
相关论文
共 19 条
  • [1] [Anonymous], 2017, DATA MINING, DOI DOI 10.1016/B978-0-12-804291-5.00014-3
  • [2] Asahara M, 2011, PEDESTRIAN MOVEMENT, P25
  • [3] Boukhechba M, 2015, 8 ACM INT C PERV TEC
  • [4] Energy Optimization for Outdoor Activity Recognition
    Boukhechba, Mehdi
    Bouzouane, Abdenour
    Bouchard, Bruno
    Gouin-Vallerand, Charles
    Giroux, Sylvain
    [J]. JOURNAL OF SENSORS, 2016, 2016
  • [5] Ezeife C.I., 2002, Advances in Artificial Intelligence, V2338, P147
  • [6] Gai YL, 2012, ADV MAT RES, V546547, P497, DOI [10.4028/www.scientific.net/AMR.546-547.497, DOI 10.4028/WWW.SCIENTIFIC.NET/AMR.546-547.497]
  • [7] Gama J, 2004, LECT NOTES ARTIF INT, V3171, P286
  • [8] Gama J., 2009, ISSUES EVALUATION ST, P329, DOI [DOI 10.1145/1557019.1557060, 10.1145/1557019.1557060]
  • [9] Gambs S., 2012, Next place prediction using mobility Markov chains, P1
  • [10] Hipp Jochen, 2000, ACM SIGKDD Explorations, V2, P58, DOI [10.1145/360402.360421, DOI 10.1145/360402.360421]