Skip navigation

DSpace

機構典藏 DSpace 系統致力於保存各式數位資料(如:文字、圖片、PDF)並使其易於取用。

點此認識 DSpace
DSpace logo
English
中文
  • 瀏覽論文
    • 校院系所
    • 出版年
    • 作者
    • 標題
    • 關鍵字
    • 指導教授
  • 搜尋 TDR
  • 授權 Q&A
    • 我的頁面
    • 接受 E-mail 通知
    • 編輯個人資料
  1. NTU Theses and Dissertations Repository
  2. 公共衛生學院
  3. 流行病學與預防醫學研究所
請用此 Handle URI 來引用此文件: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/99856
完整後設資料紀錄
DC 欄位值語言
dc.contributor.advisor王彥雯zh_TW
dc.contributor.advisorCharlotte Wangen
dc.contributor.author許國昌zh_TW
dc.contributor.authorKuo-Chang Hsuen
dc.date.accessioned2025-09-19T16:06:30Z-
dc.date.available2025-09-20-
dc.date.copyright2025-09-19-
dc.date.issued2025-
dc.date.submitted2025-08-05-
dc.identifier.citationBai, J., Di, C., Xiao, L., Evenson, K. R., LaCroix, A. Z., Crainiceanu, C. M., & Buchner, D. M. (2016). An activity index for raw accelerometry data and its comparison with other activity metrics. Plos One, 11(8), e0160644.
Berrar, D. (2019). Cross-validation. Tokyo Institute of Technology.
Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785-794). ACM.
Chung, W. Y., Purwar, A., & Sharma, A. (2008, August). Frequency domain approach for activity classification using accelerometer. In 2008 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (pp. 1120-1123). IEEE.
Crouter, S. E., Churilla, J. R., & Bassett, D. R. (2006). Estimating energy expenditure using accelerometers. European Journal of Applied Physiology, 98, 601-612.
Cruz, L., Lucio, D., & Velho, L. (2012, August). Kinect and rgbd images: Challenges and applications. In 2012 25th SIBGRAPI Conference on Graphics, Patterns and Images Tutorials (pp. 36-49). IEEE.
De Boor, C. (1972). On calculating with B-splines. Journal of Approximation Theory, 6(1), 50-62.
Fridolfsson, J., Börjesson, M., Buck, C., Ekblom, Ö., Ekblom-Bak, E., Hunsberger, M., ... & Arvidsson, D. (2019). Effects of frequency filtering on intensity and noise in accelerometer-based physical activity measurements. Sensors, 19(9), 2186.
Gjoreski, H., Lustrek, M., & Gams, M. (2011, July). Accelerometer placement for posture recognition and fall detection. In 2011 Seventh International Conference on Intelligent Environments (pp. 47-54). IEEE.
Hilden, P. (2021). Analysis Approaches for Wearable Device Data. Columbia University.
Iqbal, S. M., Mahgoub, I., Du, E., Leavitt, M. A., & Asghar, W. (2021). Advances in healthcare wearable devices. NPJ Flexible Electronics, 5(1), 9.
Kamišalić, A., Fister Jr, I., Turkanović, M., & Karakatič, S. (2018). Sensors and functionalities of non-invasive wrist-wearable devices: A review. Sensors, 18(6), 1714.
Liu, H., & Wang, L. (2018). Gesture recognition for human-robot collaboration: A review. International Journal of Industrial Ergonomics, 68, 355-367.
Liu, M. K., Lin, Y. T., Qiu, Z. W., Kuo, C. K., & Wu, C. K. (2020). Hand gesture recognition by a MMG-based wearable device. IEEE Sensors Journal, 20(24), 14703-14712.
Mardonova, M., & Choi, Y. (2018). Review of wearable device technology and its applications to the mining industry. Energies, 11(3), 547.
Nielsen, D. (2016). Tree boosting with xgboost-why does xgboost win" every" machine learning competition? Norwegian University of Science and Technology.
Patsadu, O., Nukoolkit, C., & Watanapa, B. (2012, May). Human gesture recognition using Kinect camera. In 2012 Ninth International Conference on Computer Science and Software Engineering (JCSSE) (pp. 28-32). IEEE.
Preece, S. J., Goulermas, J. Y., Kenney, L. P., & Howard, D. (2008). A comparison of feature extraction methods for the classification of dynamic activities from accelerometer data. IEEE Transactions on Biomedical Engineering, 56(3), 871-879.
Ramsay, J. O., & Silverman, B. W. (2005). Functional data analysis (2nd ed.). Springer.
Ramsay, J. O., Hooker, G., & Graves, S. (2009). Functional Data Analysis with R and MATLAB. Springer.
Razdan, A. (1999). Knot placement for B-spline curve approximation. Tempe, AZ: Arizona State University, 1(7), 8.
Seneviratne, S., Hu, Y., Nguyen, T., Lan, G., Khalifa, S., Thilakarathna, K., ... & Seneviratne, A. (2017). A survey of wearable devices and challenges. IEEE Communications Surveys & Tutorials, 19(4), 2573-2620.
Strain, T., Wijndaele, K., Dempsey, P. C., Sharp, S. J., Pearce, M., Jeon, J., ... & Brage, S. (2020). Wearable-device-measured physical activity and future health risk. Nature Medicine, 26(9), 1385-1391.
Unser, M., Aldroubi, A., & Eden, M. (1993). B-spline signal processing. I. Theory. IEEE Transactions on Signal Processing, 41(2), 821-833.
Unser, M., Aldroubi, A., & Eden, M. (1993). B-spline signal processing. II. Efficiency design and applications. IEEE Transactions on Signal Processing, 41(2), 834-848.
Van Hees, V. T., Gorzelniak, L., Dean León, E. C., Eder, M., Pias, M., Taherian, S., ... & Brage, S. (2013). Separating movement and gravity components in an acceleration signal and implications for the assessment of human daily physical activity. Plos One, 8(4), e61691.
Van Hees, V. T., Sabia, S., Anderson, K. N., Denton, S. J., Oliver, J., Catt, M., ... & Singh-Manoux, A. (2015). A novel, open access method to assess sleep duration using a wrist-worn accelerometer. Plos One, 10(11), e0142533.
Vuletic, T., Duffy, A., Hay, L., McTeague, C., Campbell, G., & Grealy, M. (2019). Systematic literature review of hand gestures used in human computer interaction interfaces. International Journal of Human-Computer Studies, 129, 74-94.
Wang, J. L., Chiou, J. M., & Müller, H. G. (2016). Functional data analysis. Annual Review of Statistics and its Application, 3, 257-295.
Wold, S. (1974). Spline functions in data analysis. Technometrics, 16(1), 1-11.
Xing, K., Ding, Z., Jiang, S., Ma, X., Yang, K., Yang, C., ... & Jiang, F. (2018, June). Hand gesture recognition based on deep learning method. In 2018 IEEE Third International Conference on Data Science in Cyberspace (DSC) (pp. 542-546). IEEE.
Yang, C. C., & Hsu, Y. L. (2010). A review of accelerometry-based wearable motion detectors for physical activity monitoring. Sensors, 10(8), 7772-7788.
-
dc.identifier.urihttp://tdr.lib.ntu.edu.tw/jspui/handle/123456789/99856-
dc.description.abstract隨著科技的進步,穿戴式裝置在醫療健康監測和人體活動識別領域的應用越來越廣泛。這些裝置通過高精度的感測器連續收集資料,提供了研究人體動作細微變化的新可能性。本研究利用函數型資料分析(Functional Data Analysis, FDA)方法,對穿戴於手腕上的裝置所收集的加速度計和陀螺儀資料進行深入分析,旨在應用函數型資料分析方法建立手勢識別模型。本研究在函數型資料分析的架構下,將資料資料經過初步處理後,使用B-Spline基底把離散型時間序列資料轉換為函數型資料,以提高其質量並降低背景噪聲;接著,應用函數型主成分分析(Functional Principal Component Analysis, FPCA)來降低資料維度,從而提取出最具代表性的特徵;最後利用提取出的主成份分數(Principal Component Score, PC Score)作為特徵變數,並透過XGBoost建立手勢辨識的分類模型。此模型應用於某公司提供的48名參與者的10種不同手勢的時間序列資料資料上,並透過交叉驗證的方式進行模型訓練與模型預測表現的評估。分析結果顯示,此流程能夠建立一個在測試集上達到接近平均80%敏感度的模型,具有不錯的分類能力和泛化能力,證明了結合FDA和FPCA的方法不僅能有效降低資料維度,還能保留用於手勢識別的關鍵動作特徵。未來可探索其他分類方法或優化FDA前處理流程,以進一步提升分類預測效果,也可以將此方法擴展到更多類型的動作識別應用中,以實現在更廣泛場景下的應用。zh_TW
dc.description.abstractWith technological advancements, wearable devices are increasingly used in health monitoring and activity recognition. These devices collect time-series data through high-precision sensors, offering new possibilities for studying subtle human movements. This study applied functional data analysis (FDA) to analyze accelerometer and gyroscope data from wrist-worn devices and establish a gesture recognition model. After data processing, B-Spline bases were used to convert discrete time series data into functional data to improve quality and reduce noise. Functional principal component analysis (FPCA) is applied to reduce dimensionality, extracting representative features. The extracted principal component (PC) scores were used as feature variables, and XGBoost was employed to build the classification model. This model is applied to the time series data of ten gestures from 48 participants. The classification model was trained, and prediction performance was evaluated through cross-validation. The results show an average sensitivity of about 80% on the testing set, demonstrating good classification and generalization abilities. This confirms that combining FDA and FPCA effectively reduces data dimensionality while preserving key features for gesture recognition. Future work can explore other classification methods or optimize the FDA preprocessing workflow to enhance prediction performance. This method can also be extended to other action recognition applications.en
dc.description.provenanceSubmitted by admin ntu (admin@lib.ntu.edu.tw) on 2025-09-19T16:06:29Z
No. of bitstreams: 0
en
dc.description.provenanceMade available in DSpace on 2025-09-19T16:06:30Z (GMT). No. of bitstreams: 0en
dc.description.tableofcontents致謝 i
中文摘要 ii
Abstract iii
目次 iv
第一章 背景與動機目的 1
1.1 穿戴式裝置介紹 1
1.2 文獻回顧 2
1.3 研究動機與目的 5
第二章 材料與方法 7
2.1 函數型資料分析 7
2.1.1 函數型資料介紹 7
2.1.2 函數型主成份分析 9
2.2 研究資料介紹 10
2.2.1 手勢資料集介紹 10
2.2.2 資料集前處理流程及函數型資料轉換 12
2.3 分類模型建立 15
2.3.1 訓練集與測試集切分 15
2.3.2 測試集函數型資料轉換及FPCA 16
2.3.3 分類器建構與評估指標 17
第三章 結果 20
3.1 探索性資料分析 20
3.1.1 手勢訊號初探 20
3.1.2 節點選擇 27
3.1.3 轉換結果 32
3.2 分類模型結果 35
3.2.1 資料集切分組結果 35
3.2.2 資料降低維度 37
3.2.3 模型表現結果 43
第四章 結論與討論 49
4.1 結論 49
4.2 討論 50
參考文獻 52
附錄 55
-
dc.language.isozh_TW-
dc.subject函數型資料分析zh_TW
dc.subject穿戴式裝置分析zh_TW
dc.subjectXGBoostzh_TW
dc.subject手勢識別zh_TW
dc.subject函數型主成份分析zh_TW
dc.subjectFunctional Principal Component Analysisen
dc.subjectGesture Recognitionen
dc.subjectXGBoosten
dc.subjectFunctional Data Analysisen
dc.subjectWearable Device Analysisen
dc.title透過函數型資料分析建構手勢辨識模型zh_TW
dc.titleConstructing a Hand Gesture Recognition Model through Functional Data Analysisen
dc.typeThesis-
dc.date.schoolyear113-2-
dc.description.degree碩士-
dc.contributor.oralexamcommittee李百靈;蕭朱杏;李美賢zh_TW
dc.contributor.oralexamcommitteePai-Ling Li;Chuhsing Kate Hsiao;Mei-Hsien Leeen
dc.subject.keyword穿戴式裝置分析,函數型資料分析,函數型主成份分析,手勢識別,XGBoost,zh_TW
dc.subject.keywordWearable Device Analysis,Functional Data Analysis,Functional Principal Component Analysis,Gesture Recognition,XGBoost,en
dc.relation.page64-
dc.identifier.doi10.6342/NTU202503895-
dc.rights.note未授權-
dc.date.accepted2025-08-06-
dc.contributor.author-college公共衛生學院-
dc.contributor.author-dept流行病學與預防醫學研究所-
dc.date.embargo-liftN/A-
顯示於系所單位:流行病學與預防醫學研究所

文件中的檔案:
檔案 大小格式 
ntu-113-2.pdf
  未授權公開取用
12.86 MBAdobe PDF
顯示文件簡單紀錄


系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。

社群連結
聯絡資訊
10617臺北市大安區羅斯福路四段1號
No.1 Sec.4, Roosevelt Rd., Taipei, Taiwan, R.O.C. 106
Tel: (02)33662353
Email: ntuetds@ntu.edu.tw
意見箱
相關連結
館藏目錄
國內圖書館整合查詢 MetaCat
臺大學術典藏 NTU Scholars
臺大圖書館數位典藏館
本站聲明
© NTU Library All Rights Reserved