提前偵測在醫學領域是一項熱門話題，因為提前偵測出疾病或臨床惡化並及早介入治療可以大幅降低患者的死亡率。許多醫學研究都深入研究不同疾病的提前偵測，包括癌症、心臟疾病甚至是近年流行的COVID-19。目前在醫院病房裡，要判斷病人是否有惡化的情況，還是要透過護理人員依據專業知識和經驗檢查病人的生命體徵或臨床情況。護理人員不足一直是醫院病房長期問題，也會降低對患者進行重新評估的頻率。不足的重新評估頻率可能會導致護理人員沒有注意到一些重要的生命體徵變化並錯過黃金治療時間。我們提出了一種基於穿戴裝置的提前偵測框架，以解決醫院人力不足而導致重新評估頻率不足的問題。穿戴裝置常見問題是它的數據可靠性，這可能是由於傳感器損壞或數據傳輸造成的。我們提出的方法能處理數據可靠性，以防止我們的模型受到不可靠數據的影響。對於提前偵測，我們提出了一個基於shapelet的概念，以在時間序列數據中找到重要趨勢，並通過觀察到shapelet的轉換構建知識圖譜。我們希望該知識圖譜能夠學習shapelet轉換之間的關係。我們假設shapelet轉換和數據可靠性是分析時間序列數據的重要組成部分。因此，我們進行消融實驗來證明我們的假設。在實驗研究中，我們提出的CST-KGE優於比較的模型，這證明了上述假設確實能讓模型表現提升。

Early detection is a popular topic in the medical domain since the discovery of disease or clinical deterioration in the early stage can help to execute rapid medical intervention, leading to a decrease in patient hospital mortality. Research has dug into different applications of early detection in cancers, critical heart disease, and COVID-19. Nowadays in hospital wards, to figure out if a patient is in a deterioration situation, the nurse should frequently check the patient's vital signs or clinical conditions based on their professional and experiences. The shortage of nurses has been a long-term problem in hospital wards and will lower the reassessment frequency for patients. Lack of reassessments may cause nurses to ignore some critical vital sign changes and to miss the golden treatment time. In this thesis, we propose an early detection framework based on a wearable device to overcome the shortage of hospital manpower caused by insufficient reassessment problems. A common problem of wearable devices is their data reliability which may be caused by error-prone nature due to body movement or data transmission. Our work will deal with data reliability to enhance our model in the error-prone data. In this thesis, we also propose a shapelet-based concept to find important patterns in time series data and build a knowledge graph by observed shapelet transitions. We expect that the graph can learn the relations between transitions of shapelet patterns. We assume that shapelet transitions and data reliability are important components in analyzing time series data. Hence, we make ablation experiments to prove our assumption. In the experimental studies, our proposed CST-KGE outperforms all baselines which proves the aforementioned assumptions can improve the model's performance.

[1] J. D. Schiffman, P. G. Fisher, and P. Gibbs, “Early detection of cancer: past, present, and future,” American Society of Clinical Oncology Educational Book, vol. 35, no. 1, pp. 57–65, 2015.
[2] V. G. Almeida and I. T. Nabney, “Early warnings of heart rate deterioration,” in 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 2016, pp. 940–943.
[3] T. Mishra, M. Wang, A. A. Metwally, G. K. Bogu, A. W. Brooks, A. Bahmani, A. Alavi, A. Celli, E. Higgs, O. Dagan-Rosenfeld et al., “Pre-symptomatic detection of covid-19 from smartwatch data,” Nature biomedical engineering, vol. 4, no. 12, pp. 1208–1220, 2020.
[4] A. Gupta, H. P. Gupta, B. Biswas, and T. Dutta, “Approaches and applications of early classification of time series: A review,” IEEE Transactions on Artificial Intelligence, vol. 1, no. 1, pp. 47–61, 2020.
[5] D. Jones, I. Mitchell, K. Hillman, and D. Story, “Defining clinical deterioration,” Resus- citation, vol. 84, no. 8, pp. 1029–1034, 2013.
[6] J. Patterson, “The effects of nurse to patient ratios.” Nursing times, vol. 107, no. 2, pp. 22–25, 2011.
[7] M. Rassin and D. Silner, “Trends in nursing staff allocation: the nurse-to-patient ratio and skill mix issues in israel,” International nursing review, vol. 54, no. 1, pp. 63–69, 2007.
[8] L. H. Aiken, D. M. Sloane, J. P. Cimiotti, S. P. Clarke, L. Flynn, J. A. Seago, J. Spetz, and H. L. Smith, “Implications of the california nurse staffing mandate for other states,” Health services research, vol. 45, no. 4, pp. 904–921, 2010.
[9] S. Shin, J. D. Park, and J. H. Shin, “Improvement plan of nurse staffing standards in korea,” Asian Nursing Research, vol. 14, no. 2, pp. 57–65, 2020.
[10] K.-F. Wu, J.-L. Hu, and H. Chiou, “Degrees of shortage and uncovered ratios for long-term care in taiwan's regions: Evidence from dynamic dea,” International journal of environmental research and public health, vol. 18, no. 2, p. 605, 2021.
[11] M. A. Rose, L. A. Hanna, S. A. Nur, and C. M. Johnson, “Utilization of electronic modified early warning score to engage rapid response team early in clinical deterioration,” Journal for nurses in professional development, vol. 31, no. 3, pp. E1–E7, 2015.
[12] J. Ludikhuize, M. Borgert, J. Binnekade, C. Subbe, D. Dongelmans, and A. Goossens, “Standardized measurement of the modified early warning score results in enhanced implementation of a rapid response system: a quasi-experimental study,” Resuscitation, vol. 85, no. 5, pp. 676–682, 2014.
[13] L. Ennis, “Paediatric early warning scores on a children’s ward: a quality improvement initiative,” Nursing children and young people, vol. 26, no. 7, 2014.
[14] R. Joshi and I. C. Gyllensten, “Changes in daily measures of blood pressure and heart rate improve weight-based detection of heart failure deterioration in patients on telemonitor- ing,” IEEE Journal of Biomedical and Health Informatics, vol. 23, no. 3, pp. 1041–1048, 2018.
[15] E. I. Bah, M.-C. Lamah, T. Fletcher, S. T. Jacob, D. M. Brett-Major, A. A. Sall, N. Shindo,
W. A. Fischer, F. Lamontagne, S. M. Saliou et al., “Clinical presentation of patients with ebola virus disease in conakry, guinea,” New England Journal of Medicine, vol. 372, no. 1, pp. 40–47, 2015.
[16] M. P. Griffin and J. R. Moorman, “Toward the early diagnosis of neonatal sepsis and sepsis-like illness using novel heart rate analysis,” Pediatrics, vol. 107, no. 1, pp. 97–104, 2001.
[17] L. Ye and E. Keogh, “Time series shapelets: a new primitive for data mining,” in Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining, 2009, pp. 947–956.
[18] S. Cho, I. Ensari, C. Weng, M. G. Kahn, K. Natarajan et al., “Factors affecting the quality of person-generated wearable device data and associated challenges: Rapid systematic review,” JMIR mHealth and uHealth, vol. 9, no. 3, p. e20738, 2021.
[19] S. K. Sharma and R. Rani, “Nurse-to-patient ratio and nurse staffing norms for hospitals in india: a critical analysis of national benchmarks,” Journal of family medicine and primary care, vol. 9, no. 6, p. 2631, 2020.
[20] L. H. Aiken, S. P. Clarke, D. M. Sloane, J. Sochalski, and J. H. Silber, “Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction,” Jama, vol. 288, no. 16, pp. 1987–1993, 2002.
[21] T. L. McGhee, P. Weaver, S. Solo, and M. Hobbs, “Vital signs reassessment frequency recommendation,” Nursing Management, vol. 47, no. 9, pp. 11–12, 2016.
[22] L. Rose and S. P. Clarke, “Vital signs,” AJN The American Journal of Nursing, vol. 110, no. 5, p. 11, 2010.
[23] S. Seneviratne, Y. Hu, T. Nguyen, G. Lan, S. Khalifa, K. Thilakarathna, M. Hassan, and A. Seneviratne, “A survey of wearable devices and challenges,” IEEE Communications Surveys & Tutorials, vol. 19, no. 4, pp. 2573–2620, 2017.
[24] A. Pantelopoulos and N. G. Bourbakis, “A survey on wearable sensor-based systems for health monitoring and prognosis,” IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), vol. 40, no. 1, pp. 1–12, 2009.
[25] S. Beniczky, S. Wiebe, J. Jeppesen, W. O. Tatum, M. Brazdil, Y. Wang, S. T. Herman, and P. Ryvlin, “Automated seizure detection using wearable devices: A clinical practice guideline of the international league against epilepsy and the international federation of clinical neurophysiology,” Clinical Neurophysiology, vol. 132, no. 5, pp. 1173–1184, 2021.
[26] M. Nasseri, T. Pal Attia, B. Joseph, N. M. Gregg, E. S. Nurse, P. F. Viana, G. Worrell,
M. Du¨mpelmann, M. P. Richardson, D. R. Freestone et al., “Ambulatory seizure forecasting with a wrist-worn device using long-short term memory deep learning,” Scientific reports, vol. 11, no. 1, pp. 1–9, 2021.
[27] X. Shi, Z. Chen, H. Wang, D.-Y. Yeung, W.-K. Wong, and W.-c. Woo, “Convolutional lstm network: A machine learning approach for precipitation nowcasting,” Advances in neural information processing systems, vol. 28, 2015.
[28] H. Yao, X. Tang, H. Wei, G. Zheng, and Z. Li, “Revisiting spatial-temporal similarity: A deep learning framework for traffic prediction,” in Proceedings of the AAAI conference on artificial intelligence, vol. 33, no. 01, 2019, pp. 5668–5675.
[29] C. Sun, S. Hong, M. Song, and H. Li, “A review of deep learning methods for irregularly sampled medical time series data,” arXiv preprint arXiv:2010.12493, 2020.
[30] X. Wu, B. Shi, Y. Dong, C. Huang, L. Faust, and N. V. Chawla, “Restful: Resolution-aware forecasting of behavioral time series data,” in Proceedings of the 27th ACM International Conference on Information and Knowledge Management, 2018, pp. 1073–1082.
[31] S. Song, C. Li, and X. Zhang, “Turn waste into wealth: On simultaneous clustering and cleaning over dirty data,” in Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2015, pp. 1115–1124.
[32] A. Zhang, S. Song, J. Wang, and P. S. Yu, “Time series data cleaning: From anomaly detection to anomaly repairing,” Proceedings of the VLDB Endowment, vol. 10, no. 10, pp. 1046–1057, 2017.
[33] A. Zhang, S. Song, Y. Sun, and J. Wang, “Learning individual models for imputation,” in 2019 IEEE 35th International Conference on Data Engineering (ICDE). IEEE, 2019, pp. 160–171.
[34] H.-F. Yu, N. Rao, and I. S. Dhillon, “Temporal regularized matrix factorization for high- dimensional time series prediction,” Advances in neural information processing systems, vol. 29, 2016.
[35] Z. Ghahramani and M. Jordan, “Supervised learning from incomplete data via an em approach,” Advances in neural information processing systems, vol. 6, 1993.
[36] Y. Luo, Y. Zhang, X. Cai, and X. Yuan, “E2gan: End-to-end generative adversarial network for multivariate time series imputation,” in Proceedings of the 28th international joint conference on artificial intelligence. AAAI Press, 2019, pp. 3094–3100.
[37] D. Sopic, A. Aminifar, A. Aminifar, and D. Atienza, “Real-time event-driven classification technique for early detection and prevention of myocardial infarction on wearable systems,” IEEE transactions on biomedical circuits and systems, vol. 12, no. 5, pp. 982–992, 2018.
[38] C. Sun, S. Hong, M. Song, H. Li, and Z. Wang, “Predicting covid-19 disease progression and patient outcomes based on temporal deep learning,” BMC Medical Informatics and Decision Making, vol. 21, no. 1, pp. 1–16, 2021.
[39] B. Liu, J. Li, C. Chen, W. Tan, Q. Chen, and M. Zhou, “Efficient motif discovery for large-scale time series in healthcare,” IEEE Transactions on Industrial Informatics, vol. 11, no. 3, pp. 583–590, 2015.
[40] G. Chen, G. Lu, W. Shang, and Z. Xie, “Automated change-point detection of eeg signals based on structural time-series analysis,” IEEE Access, vol. 7, pp. 180 168–180 180, 2019.
[41] K. Fauvel, D. Balouek-Thomert, D. Melgar, P. Silva, A. Simonet, G. Antoniu, A. Costan, V. Masson, M. Parashar, I. Rodero et al., “A distributed multi-sensor machine learning approach to earthquake early warning,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 01, 2020, pp. 403–411.
[42] N. Hatami and C. Chira, “Classifiers with a reject option for early time-series classification,” in 2013 IEEE symposium on computational intelligence and ensemble learning (CIEL). IEEE, 2013, pp. 9–16.
[43] G. He, Y. Duan, G. Zhou, and L. Wang, “Early classification on multivariate time series with core features,” in International Conference on Database and Expert Systems Applications. Springer, 2014, pp. 410–422.
[44] A. Dachraoui, A. Bondu, and A. Cornu´ejols, “Early classification of time series as a nonmyopic sequential decision making problem,” in Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, 2015, pp. 433–447.
[45] M. F. Ghalwash, D. Ramljak, and Z. Obradovi´c, “Early classification of multivariate time series using a hybrid hmm/svm model,” in 2012 IEEE International Conference on Bioinformatics and Biomedicine. IEEE, 2012, pp. 1–6.
[46] F. E. Shamout, T. Zhu, P. Sharma, P. J. Watkinson, and D. A. Clifton, “Deep interpretable early warning system for the detection of clinical deterioration,” IEEE journal of biomedical and health informatics, vol. 24, no. 2, pp. 437–446, 2019.
[47] Z. Xing, J. Pei, and P. S. Yu, “Early classification on time series,” Knowledge and information systems, vol. 31, no. 1, pp. 105–127, 2012.
[48] C. Ma, X. Weng, and Z. Shan, “Early classification of multivariate time series based on piecewise aggregate approximation,” in International Conference on Health Information Science. Springer, 2017, pp. 81–88.
[49] Z. Xing, J. Pei, and S. Y. Philip, “Early prediction on time series: A nearest neighbor approach,” in Twenty-First International Joint Conference on Artificial Intelligence. Citeseer, 2009.
[50] Z. Xing, J. Pei, P. S. Yu, and K. Wang, “Extracting interpretable features for early classification on time series,” in Proceedings of the 2011 SIAM international conference on data mining. SIAM, 2011, pp. 247–258.
[51] H.-S. Huang, C.-L. Liu, and V. S. Tseng, “Multivariate time series early classification using multi-domain deep neural network,” in 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA). IEEE, 2018, pp. 90–98.
[52] M. Rußwurm, S. Lefevre, N. Courty, R. Emonet, M. K¨orner, and R. Tavenard, “End-to-end learning for early classification of time series,” arXiv preprint arXiv:1901.10681, 2019.
[53] C. Martinez, G. Perrin, E. Ramasso, and M. Rombaut, “A deep reinforcement learning approach for early classification of time series,” in 2018 26th European Signal Processing Conference (EUSIPCO). IEEE, 2018, pp. 2030–2034.
[54] M. G. Pinheiro, S. C. Madeira, and A. P. Francisco, “Shapelets to classify energy demand time series,” Energies, vol. 15, no. 8, p. 2960, 2022.
[55] P. Kidger, J. Morrill, and T. Lyons, “Generalised interpretable shapelets for irregular time series,” arXiv preprint arXiv:2005.13948, 2020.
[56] X. Renard, M. Rifqi, W. Erray, and M. Detyniecki, “Random-shapelet: an algorithm for fast shapelet discovery,” in 2015 IEEE international conference on data science and advanced analytics (DSAA). IEEE, 2015, pp. 1–10.
[57] L. Hou, J. Kwok, and J. Zurada, “Efficient learning of timeseries shapelets,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 30, no. 1, 2016.
[58] Z. Fang, P. Wang, and W. Wang, “Efficient learning interpretable shapelets for accurate time series classification,” in 2018 IEEE 34th International Conference on Data Engineering (ICDE). IEEE, 2018, pp. 497–508.
[59] G. Li, B. K. K. Choi, J. Xu, S. S. Bhowmick, K.-P. Chun, and G. L. Wong, “Efficient shapelet discovery for time series classification,” IEEE Transactions on Knowledge and Data Engineering, 2020.
[60] J. Lines, L. M. Davis, J. Hills, and A. Bagnall, “A shapelet transform for time series classification,” in Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining, 2012, pp. 289–297.
[61] J. Grabocka, N. Schilling, M. Wistuba, and L. Schmidt-Thieme, “Learning time-series shapelets,” in Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining, 2014, pp. 392–401.
[62] L. Yao, Y. Li, Y. Li, H. Zhang, M. Huai, J. Gao, and A. Zhang, “Dtec: Distance trans- formation based early time series classification,” in Proceedings of the 2019 SIAM International Conference on Data Mining. SIAM, 2019, pp. 486–494.
[63] M. F. Ghalwash, V. Radosavljevic, and Z. Obradovic, “Extraction of interpretable multivariate patterns for early diagnostics,” in 2013 IEEE 13th International Conference on Data Mining. IEEE, 2013, pp. 201–210.
[64] G. He, Y. Duan, R. Peng, X. Jing, T. Qian, and L. Wang, “Early classification on multivariate time series,” Neurocomputing, vol. 149, pp. 777–787, 2015.
[65] Z. Cheng, Y. Yang, W. Wang, W. Hu, Y. Zhuang, and G. Song, “Time2graph: Revisiting time series modeling with dynamic shapelets,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 04, 2020, pp. 3617–3624.
[66] W. Hu, Y. Yang, Z. Cheng, C. Yang, and X. Ren, “Time-series event prediction with evolutionary state graph,” in Proceedings of the 14th ACM International Conference on Web Search and Data Mining, 2021, pp. 580–588.
[67] X. Yang, Y. Zheng, Y. Zhang, D. S.-H. Wong, and W. Yang, “Bearing remaining useful life prediction based on regression shapalet and graph neural network,” IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1–12, 2022.
[68] Y. Liu, X. Shi, L. Pierce, and X. Ren, “Characterizing and forecasting user engagement with in-app action graph: A case study of snapchat,” in Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019, pp. 2023–2031.
[69] X. Geng, Y. Li, L. Wang, L. Zhang, Q. Yang, J. Ye, and Y. Liu, “Spatiotemporal multi- graph convolution network for ride-hailing demand forecasting,” in Proceedings of the AAAI conference on artificial intelligence, vol. 33, no. 01, 2019, pp. 3656–3663.
[70] P. Battaglia, R. Pascanu, M. Lai, D. Jimenez Rezende et al., “Interaction networks for learning about objects, relations and physics,” Advances in neural information processing systems, vol. 29, 2016.
[71] A. Sanchez-Gonzalez, N. Heess, J. T. Springenberg, J. Merel, M. Riedmiller, R. Hadsell, and P. Battaglia, “Graph networks as learnable physics engines for inference and control,” in International Conference on Machine Learning. PMLR, 2018, pp. 4470–4479.
[72] Q. Guo, F. Zhuang, C. Qin, H. Zhu, X. Xie, H. Xiong, and Q. He, “A survey on knowledge graph-based recommender systems,” IEEE Transactions on Knowledge and Data Engineering, 2020.
[73] A. Bordes, N. Usunier, A. Garcia-Duran, J. Weston, and O. Yakhnenko, “Translating embeddings for modeling multi-relational data,” Advances in neural information processing systems, vol. 26, 2013.
[74] Z. Wang, J. Zhang, J. Feng, and Z. Chen, “Knowledge graph embedding by translating on hyperplanes,” in Proceedings of the AAAI conference on artificial intelligence, vol. 28, no. 1, 2014.
[75] M. Nickel, V. Tresp, and H.-P. Kriegel, “A three-way model for collective learning on multi-relational data,” in Icml, 2011.
[76] B. Yang, W.-t. Yih, X. He, J. Gao, and L. Deng, “Embedding entities and relations for learning and inference in knowledge bases,” arXiv preprint arXiv:1412.6575, 2014.
[77] T. Chen and C. Guestrin, “Xgboost: A scalable tree boosting system,” in Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 2016, pp. 785–794.
[78] T. Flenady, T. Dwyer, and J. Applegarth, “Accurate respiratory rates count: So should you!” Australasian emergency nursing journal, vol. 20, no. 1, pp. 45–47, 2017.