Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model

Xinze Guan, Raviv Raich, Weng-Keen Wong
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:2330-2339, 2016.

Abstract

Activity recognition from sensor data has spurred a great deal of interest due to its impact on health care. Prior work on activity recognition from multivariate time series data has mainly applied supervised learning techniques which require a high degree of annotation effort to produce training data with the start and end times of each activity. In order to reduce the annotation effort, we present a weakly supervised approach based on multi-instance learning. We introduce a generative graphical model for multi-instance learning on time series data based on an auto-regressive hidden Markov model. Our model has a number of advantages, including the ability to produce both bag and instance-level predictions as well as an efficient exact inference algorithm based on dynamic programming.

Cite this Paper

BibTeX
@InProceedings{pmlr-v48-guan16, title = {Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model}, author = {Guan, Xinze and Raich, Raviv and Wong, Weng-Keen}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {2330--2339}, year = {2016}, editor = {Balcan, Maria Florina and Weinberger, Kilian Q.}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/guan16.pdf}, url = {https://proceedings.mlr.press/v48/guan16.html}, abstract = {Activity recognition from sensor data has spurred a great deal of interest due to its impact on health care. Prior work on activity recognition from multivariate time series data has mainly applied supervised learning techniques which require a high degree of annotation effort to produce training data with the start and end times of each activity. In order to reduce the annotation effort, we present a weakly supervised approach based on multi-instance learning. We introduce a generative graphical model for multi-instance learning on time series data based on an auto-regressive hidden Markov model. Our model has a number of advantages, including the ability to produce both bag and instance-level predictions as well as an efficient exact inference algorithm based on dynamic programming.} }
Endnote
%0 Conference Paper %T Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model %A Xinze Guan %A Raviv Raich %A Weng-Keen Wong %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-guan16 %I PMLR %P 2330--2339 %U https://proceedings.mlr.press/v48/guan16.html %V 48 %X Activity recognition from sensor data has spurred a great deal of interest due to its impact on health care. Prior work on activity recognition from multivariate time series data has mainly applied supervised learning techniques which require a high degree of annotation effort to produce training data with the start and end times of each activity. In order to reduce the annotation effort, we present a weakly supervised approach based on multi-instance learning. We introduce a generative graphical model for multi-instance learning on time series data based on an auto-regressive hidden Markov model. Our model has a number of advantages, including the ability to produce both bag and instance-level predictions as well as an efficient exact inference algorithm based on dynamic programming.
RIS
TY - CPAPER TI - Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model AU - Xinze Guan AU - Raviv Raich AU - Weng-Keen Wong BT - Proceedings of The 33rd International Conference on Machine Learning DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-guan16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 2330 EP - 2339 L1 - http://proceedings.mlr.press/v48/guan16.pdf UR - https://proceedings.mlr.press/v48/guan16.html AB - Activity recognition from sensor data has spurred a great deal of interest due to its impact on health care. Prior work on activity recognition from multivariate time series data has mainly applied supervised learning techniques which require a high degree of annotation effort to produce training data with the start and end times of each activity. In order to reduce the annotation effort, we present a weakly supervised approach based on multi-instance learning. We introduce a generative graphical model for multi-instance learning on time series data based on an auto-regressive hidden Markov model. Our model has a number of advantages, including the ability to produce both bag and instance-level predictions as well as an efficient exact inference algorithm based on dynamic programming. ER -
APA
Guan, X., Raich, R. & Wong, W.. (2016). Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:2330-2339 Available from https://proceedings.mlr.press/v48/guan16.html.

Related Material