基于流式感知数据的行为识别Behavior Recognition Based on Sensory Sata Stream
朱平飞,卢耀宗,罗艺闯,强劲
摘要(Abstract)：
行为识别侧重于通过使用传感器数据来推断当前用户的活动,典型的行为识别技术通常基于点对点的方法来处理传感器感知到的数据,其中,监督学习算法在该领域有广泛的应用。本文提出了一种基于聚类的分类算法来进行行为识别,该算法采用增量式学习来挖掘数据流中的用户行为,通过将不同的活动行为赋予不同的类,融合监督、无监督和主动学习算法,并结合混合相似性度量方法建立一个鲁棒的识别系统。
关键词(KeyWords)： 行为识别;流式数据;主动学习;增量学习;混合相似性度量
基金项目(Foundation):
作者(Author): 朱平飞,卢耀宗,罗艺闯,强劲
DOI: 10.16453/j.issn.2095-8595.2014.01.019
参考文献(References)：
- [1]S.J.Preece,J.Y.Goulermas,L.P.J.Kenney,and etc.“Activity identification using body-mounted sensors:a review of classification techniques[J],”Physiological Measurement,vol.30,no.4,pp.R1–R33,2009.
- [2]J.R.Kwapisz,G.M.Weiss,and S.A.Moore.“Activity recognition using cell phone accelerometers[J],”SIGKDD Explor.,vol.12,pp.74–82,2011.
- [3]I.H.Witten,E.Frank,and Mark Hall.Data Mining:Practical Machine Learning Tools and Techniques[M].Morgan Kaufmann Publishers In,Morgan Kaufmann,2011.
- [4]M.Helmi and S.M.T.Almodarresi.“Human activity recognition using a fuzzy inference system[J],”in IEEE International Conference on Fuzzy Systems,2009,pp.1897–1902.
- [5]M.hee Lee,J.Kim,K.Kim,and etc.,“Physical activity recognition using a single tri-axis accelerometer[J],”in Proceedings of WCECS’09,2009,pp.14–17.
- [6]B.Longstaff,S.Reddy,and D.Estrin.“Improving activity classification for health applications on mobile devices using active and semi-supervised learning[J],”in Pervasive Computing Technologies for Healthcare(PervasiveH ealth)2010,march 2010,pp.1–7.
- [7]ShonaliK rishnaswamy,Bala Srinivasan.“Mobile Activity Recognition Using Contextual Reasoning and Ubiquitous Data Stream Processing[J]”.Australasian Computing Doctoral Consortium 2012.
- [8]E.J.Spinosa,A.P.de Leon F.de Carvalho,and J.a.Gama.“Olindda:a cluster-based approach for detecting novelty and concept drift in data streams[J],”in Proceedings of the 2007 ACM symposium on Applied computing.2007,pp.448–452.