臺灣博碩士論文加值系統

由於醫療技術的進步，人類的致死率獲得一定程度的降低，而這也代表著人口老化社會的來臨。歸功於現在無線感測網路與物聯網的發展，將可以隨時隨地監控年長者的日常生活活動，且長期的日常生活資料將可以被分析，進而建立出長者的行為模型。概括而言，行為模式是關於長者生活行為非常重要之資訊，有了行為模型的幫助，照護者可以得知長者的生活習慣或習慣的改變，並且給予即時且適當的幫助。目前為止，已經有非常多的研究有關於如何學習人類的行為。然而，他們的假設可能過於簡單或者不能全面的將現實中複雜的人類行為萃取成有用的資訊。因此，我們提出一個以日常生活模式感知的行為學習框架，此框架著重於探勘年長者的日常行為跟預測，此框架發掘諸如一個活動的發生時間以及長度等週期性資訊，還有活動與活動之間的關聯性分析，以週期性跟循序性兩種不同方式學習長者的行為模式。透過多種行為模式的學習，將可以彌補單一模式的不足，達到準確的活動預測與異常行為的初步判讀。長者可以藉此了解自己的生活模式，照護者同樣可以得到有幫助的資訊。當發生異常時，我們的系統將及時給予照護者與長者本人通知。我們使用兩個公開的資料集與我們自己的資料集進行活動預測的評估，另外使用我們的資料集做異常偵測的評估。

Due to longer life expectancy and declining fertility rates, ageing society is arriving. Nowadays, close monitoring of daily activities of elders is enabled by employment of the advanced wireless sensor networks and Internet of Things, whose large quantity da-ta are then analyzed by activity recognition techniques whereby their behavior can be accurately modelled. In general, behavior is important information about how elders live, and caregivers can thus take care of elders more easily with the help from that infor-mation. So far, there are many research results related to learning of human behavior; however, their assumptions are usually either too simple or inflexible to account for complex human behavior in real life, which change dynamically depending on daily life pattern. We here present a daily pattern based framework for human behavior learning and prediction. Such framework discovers contexts like start time and duration of activ-ity from resident’s real life data and relations between activities, respectively. Learning multiple behavior models can achieve the accurate activity prediction and the prelimi-nary anomaly detection. Elders can examine the lifestyle of themselves and caregivers also can obtain helpful information. When anomaly happens, the system notifies elders and caregivers immediately. We evaluate the prediction accuracy on two public datasets and our own datasets, and also the anomaly detection rate on our own data, and the ex-perimental results have shown promising results.

中文摘要 i
ABSTRACT ii
Table of Contents iii
List of Figures vi
List of Tables vii
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Challenges 3
1.2.1 Dealing with the complex human behavior with inclusion of temporal variability 3
1.2.2 Adaptive learning of human behavior model 5
1.3 Related Work 5
1.3.1 Statistics and cluster based methods 6
1.3.2 Holmes 7
1.3.3 CRAFFT 8
1.3.4 PreHeat 8
1.3.5 C. –L, Wu et al. 8
1.3.6 Soulas et al. 9
1.4 Objective 9
1.4.1 Propose a Human Behavior Learning Framework 10
1.4.2 Adaptive the behavior model and make them dynamic 12
1.4.3 Provide a friendly health support mechanism for caregivers 13
1.5 System Overview 14
1.6 Thesis Organization 15
Chapter 2 Preliminaries 17
2.1 Internet of Things and Pervasive Environment 17
2.2 Temporal Features 19
2.2.1 Gaussian Mixture Model 19
2.2.2 Sequential Discounting Laplace Estimation 21
2.2.3 Gradient Boosting Regression Tree 23
2.2.4 Kullback–Leibler Divergence 25
2.3 Sequential Relation 25
2.3.1 Bayesian Network 26
Chapter 3 Human Behavior Learning 29
3.1 Date Preprocessing 29
3.2 Learning Framework Overview 31
3.2.1 Estimation of Per-day Activity Distribution 32
3.2.2 Daily Life Pattern-based Day-by-Day Merge 34
3.2.3 Daily Life Pattern-based Periodical Pattern Modelling 36
3.2.4 Activity Relation Construction 40
Chapter 4 IoT Based Health Support System 45
4.1 System Overview 45
4.2 Online Phase 47
4.2.1 Prediction Phase 47
4.2.2 Anomaly Detection Phase 49
4.3 Adaptive Learning 51
4.3.1 Adaptation of Start Time Group 52
4.3.2 Adaptation of Start Time Group 53
4.3.3 Adaptation of Sequential Relation of Activities 54
Chapter 5 System Evaluation 56
5.1 Experimental Environment 56
5.2 Evaluation of Human Behavior Learning 58
5.2.1 Discussion of Daily Life Pattern 58
5.2.2 Evaluation of Periodical Pattern 61
5.2.3 Evaluation of Activity Prediction 66
5.2.4 Evaluation of Preliminary Anomaly Detection 69
Chapter 6 Conclusion 72
6.1 Summary 72
6.2 Future Work 73
REFERENCE 75

[1]W. H. Organization, "Global health and aging," Bethesda: National Institutes of Health, 2011.
[2]H. Banaee, M. U. Ahmed, and A. Loutfi, "Data mining for wearable sensors in health monitoring systems: a review of recent trends and challenges," Sensors, vol. 13, pp. 17472-17500, 2013.
[3]A. Pantelopoulos and N. G. Bourbakis, "A survey on wearable sensor-based systems for health monitoring and prognosis," Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on, vol. 40, pp. 1-12, 2010.
[4]L. M. Orlov, "Technology for Aging in Place."
[5]G. Tao, W. Zhanqing, T. Xianping, H. K. Pung, and L. Jian, "epSICAR: An Emerging Patterns based approach to sequential, interleaved and Concurrent Activity Recognition," in Pervasive Computing and Communications, 2009. PerCom 2009. IEEE International Conference on, 2009, pp. 1-9.
[6]K. Gayathri, S. Elias, and B. Ravindran, "Hierarchical activity recognition for dementia care using Markov Logic Network," Personal and Ubiquitous Computing, vol. 19, pp. 271-285, 2015.
[7]C.-L. Wu, T.-C. Chiang, L.-C. Fu, and Y.-C. Zeng, "Nonparametric Discovery of Contexts and Preferences in Smart Home Environments," in Systems, Man, and Cybernetics (SMC), 2015 IEEE International Conference on, 2015, pp. 2817-2822.
[8]Y.-H. Chen, M.-J. Tsai, L.-C. Fu, C.-H. Chen, C.-L. Wu, and Y.-C. Zeng, "Monitoring Elder''s Living Activity Using Ambient and Body Sensor Network in Smart Home," in Systems, Man, and Cybernetics (SMC), 2015 IEEE International Conference on, 2015, pp. 2962-2967.
[9]E. Nazerfard and D. J. Cook, "CRAFFT: an activity prediction model based on Bayesian networks," Journal of ambient intelligence and humanized computing, vol. 6, pp. 193-205, 2015.
[10]J. Soulas, P. Lenca, and A. Thépaut, "Unsupervised discovery of activities of daily living characterized by their periodicity and variability," Engineering Applications of Artificial Intelligence, vol. 45, pp. 90-102, 2015.
[11]C.-L. Wu, W.-C. Chen, Y.-S. Tseng, L.-C. Fu, and C.-H. Lu, "Anticipatory Reasoning for a Proactive Context-Aware Energy Saving System," in IEEE Internet of Things, 2014, pp. 228-234.
[12]A. Aztiria, J. C. Augusto, R. Basagoiti, A. Izaguirre, and D. J. Cook, "Learning frequent behaviors of the users in intelligent environments," Systems, Man, and Cybernetics: Systems, IEEE Transactions on, vol. 43, pp. 1265-1278, 2013.
[13]A. Kailas and M. A. Ingram, "Wireless communications technology in telehealth systems," in Wireless Communication, Vehicular Technology, Information Theory and Aerospace & Electronic Systems Technology, 2009. Wireless VITAE 2009. 1st International Conference on, 2009, pp. 926-930.
[14]H. Alemdar and C. Ersoy, "Wireless sensor networks for healthcare: A survey," Computer Networks, vol. 54, pp. 2688-2710, 2010.
[15]E. M. Tapia, S. S. Intille, and K. Larson, Activity recognition in the home using simple and ubiquitous sensors: Springer, 2004.
[16]Y. Han, M. Han, S. Lee, A. Sarkar, and Y.-K. Lee, "A framework for supervising lifestyle diseases using long-term activity monitoring," Sensors, vol. 12, pp. 5363-5379, 2012.
[17]N. Suryadevara, S. C. Mukhopadhyay, R. Wang, and R. Rayudu, "Forecasting the behavior of an elderly using wireless sensors data in a smart home," Engineering Applications of Artificial Intelligence, vol. 26, pp. 2641-2652, 2013.
[18]E. Nazerfard, P. Rashidi, and D. J. Cook, "Discovering Temporal Features and Relations of Activity Patterns," in Data Mining Workshops (ICDMW), 2010 IEEE International Conference on, 2010, pp. 1069-1075.
[19]E. Hoque, R. F. Dickerson, S. M. Preum, M. Hanson, A. Barth, and J. A. Stankovic, "Holmes: A comprehensive anomaly detection system for daily in-home activities," in Distributed Computing in Sensor Systems (DCOSS), 2015 International Conference on, 2015, pp. 40-51.
[20]J. Scott, A. J. B. Brush, J. Krumm, B. Meyers, M. Hazas, S. Hodges, et al., "PreHeat: controlling home heating using occupancy prediction," presented at the Proceedings of the 13th international conference on Ubiquitous computing, Beijing, China, 2011.
[21]A. Akl, J. Snoek, and A. Mihailidis, "Unobtrusive Detection of Mild Cognitive Impairment in Older Adults Through Home Monitoring," IEEE Journal of Biomedical and Health Informatics, vol. PP, pp. 1-1, 2015.
[22]J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, "Internet of Things (IoT): A vision, architectural elements, and future directions," Future Generation Computer Systems, vol. 29, pp. 1645-1660, 2013.
[23]G. D. Abowd, "Hot or Not?: Moving Forward from Weiser''s Vision of Ubiquitous Computing," in Proceedings of the 17th International Workshop on Mobile Computing Systems and Applications, 2016, pp. 1-1.
[24]U. Hansmann, L. Merk, M. S. Nicklous, and T. Stober, Pervasive computing: The mobile world: Springer Science & Business Media, 2003.
[25]C. H. Lu, C. L. Wu, M. Y. Weng, W. C. Chen, and L. C. Fu, "Context-Aware Energy Saving System with Multiple Comfort-Constrained Optimization in M2M-Based Home Environment," IEEE Transactions on Automation Science and Engineering, vol. PP, pp. 1-15, 2015.
[26]K.-L. Huang, Y.-H. Chen, C.-F. Liao, C. C.-H. Chen, and L.-C. Fu, "Daily health assessment system using prediction model for self-rated health by vital sign pattern," in Healthcare Innovation Conference (HIC), 2014 IEEE, 2014, pp. 95-98.
[27]C. Biernacki, G. Celeux, and G. Govaert, "Assessing a mixture model for clustering with the integrated completed likelihood," IEEE transactions on pattern analysis and machine intelligence, vol. 22, pp. 719-725, 2000.
[28]A. P. Dempster, N. M. Laird, and D. B. Rubin, "Maximum likelihood from incomplete data via the EM algorithm," Journal of the royal statistical society. Series B (methodological), pp. 1-38, 1977.
[29]K. Yamanishi, J.-I. Takeuchi, G. Williams, and P. Milne, "On-line unsupervised outlier detection using finite mixtures with discounting learning algorithms," Data Mining and Knowledge Discovery, vol. 8, pp. 275-300, 2004.
[30]T. Mori, A. Fujii, M. Shimosaka, H. Noguchi, and T. Sato, "Typical behavior patterns extraction and anomaly detection algorithm based on accumulated home sensor data," in Future Generation Communication and Networking (FGCN 2007), 2007, pp. 12-18.
[31]J. H. Friedman, "Stochastic gradient boosting," Computational Statistics & Data Analysis, vol. 38, pp. 367-378, 2002.
[32]J. Ye, J.-H. Chow, J. Chen, and Z. Zheng, "Stochastic gradient boosted distributed decision trees," in Proceedings of the 18th ACM conference on Information and knowledge management, 2009, pp. 2061-2064.
[33]S. Kullback and R. A. Leibler, "On information and sufficiency," The annals of mathematical statistics, vol. 22, pp. 79-86, 1951.
[34]K. Gopalratnam and D. J. Cook, "Online sequential prediction via incremental parsing: The active lezi algorithm," Intelligent Systems, IEEE, vol. 22, pp. 52-58, 2007.
[35]N. Friedman, D. Geiger, and M. Goldszmidt, "Bayesian network classifiers," Machine learning, vol. 29, pp. 131-163, 1997.
[36]T. Mori, R. Urushibata, M. Shimosaka, H. Noguchi, and T. Sato, "Anomaly detection algorithm based on life pattern extraction from accumulated pyroelectric sensor data," in Intelligent Robots and Systems, 2008. IROS 2008. IEEE/RSJ International Conference on, 2008, pp. 2545-2552.
[37]A. Mohan, Z. Chen, and K. Q. Weinberger, "Web-Search Ranking with Initialized Gradient Boosted Regression Trees," in Yahoo! Learning to Rank Challenge, 2011, pp. 77-89.
[38]D. J. Cook, "Learning setting-generalized activity models for smart spaces," IEEE intelligent systems, vol. 2010, p. 1, 2010.
[39]A. Akl, J. Snoek, and A. Mihailidis, "Unobtrusive detection of mild cognitive impairment in older adults through home monitoring," 2015.