Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge

Jinyu Xie, Qian Wang

    Research output: Contribution to journalConference article

    Abstract

    There is a growing trend of applying machine learning techniques in time series prediction tasks. In the meanwhile, the classic autoregression models has been widely used in time series prediction for decades. In this paper, experiments are conducted to compare the performances of multiple popular machine learning algorithms including two major types of deep learning approaches, with the classic autoregression with exogenous inputs (ARX) model on this particular Blood Glucose Level Prediction (BGLP) Challenge. We tried two types of methods to perform multi-step prediction: recursive method and direct method. The recursive method needs future input feature information. The results show there is no significant difference between the machine learning models and the classic ARX model. In fact, the ARX model achieved the lowest average Root Mean Square Error (RMSE) across subjects in the test data when recursive method was used for multi-step prediction.

    Original languageEnglish (US)
    Pages (from-to)97-102
    Number of pages6
    JournalCEUR Workshop Proceedings
    Volume2148
    StatePublished - Jan 1 2018
    Event3rd International Workshop on Knowledge Discovery in Healthcare Data, KDH@IJCAI-ECAI 2018 - Stockholm, Sweden
    Duration: Jul 13 2018 → …

    Fingerprint

    Glucose
    Learning systems
    Time series
    Blood
    Mean square error
    Learning algorithms
    Experiments

    All Science Journal Classification (ASJC) codes

    • Computer Science(all)

    Cite this

    @article{ea5fc8997e5e4b34b0556b3972833b87,
    title = "Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge",
    abstract = "There is a growing trend of applying machine learning techniques in time series prediction tasks. In the meanwhile, the classic autoregression models has been widely used in time series prediction for decades. In this paper, experiments are conducted to compare the performances of multiple popular machine learning algorithms including two major types of deep learning approaches, with the classic autoregression with exogenous inputs (ARX) model on this particular Blood Glucose Level Prediction (BGLP) Challenge. We tried two types of methods to perform multi-step prediction: recursive method and direct method. The recursive method needs future input feature information. The results show there is no significant difference between the machine learning models and the classic ARX model. In fact, the ARX model achieved the lowest average Root Mean Square Error (RMSE) across subjects in the test data when recursive method was used for multi-step prediction.",
    author = "Jinyu Xie and Qian Wang",
    year = "2018",
    month = "1",
    day = "1",
    language = "English (US)",
    volume = "2148",
    pages = "97--102",
    journal = "CEUR Workshop Proceedings",
    issn = "1613-0073",
    publisher = "CEUR-WS",

    }

    Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge. / Xie, Jinyu; Wang, Qian.

    In: CEUR Workshop Proceedings, Vol. 2148, 01.01.2018, p. 97-102.

    Research output: Contribution to journalConference article

    TY - JOUR

    T1 - Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge

    AU - Xie, Jinyu

    AU - Wang, Qian

    PY - 2018/1/1

    Y1 - 2018/1/1

    N2 - There is a growing trend of applying machine learning techniques in time series prediction tasks. In the meanwhile, the classic autoregression models has been widely used in time series prediction for decades. In this paper, experiments are conducted to compare the performances of multiple popular machine learning algorithms including two major types of deep learning approaches, with the classic autoregression with exogenous inputs (ARX) model on this particular Blood Glucose Level Prediction (BGLP) Challenge. We tried two types of methods to perform multi-step prediction: recursive method and direct method. The recursive method needs future input feature information. The results show there is no significant difference between the machine learning models and the classic ARX model. In fact, the ARX model achieved the lowest average Root Mean Square Error (RMSE) across subjects in the test data when recursive method was used for multi-step prediction.

    AB - There is a growing trend of applying machine learning techniques in time series prediction tasks. In the meanwhile, the classic autoregression models has been widely used in time series prediction for decades. In this paper, experiments are conducted to compare the performances of multiple popular machine learning algorithms including two major types of deep learning approaches, with the classic autoregression with exogenous inputs (ARX) model on this particular Blood Glucose Level Prediction (BGLP) Challenge. We tried two types of methods to perform multi-step prediction: recursive method and direct method. The recursive method needs future input feature information. The results show there is no significant difference between the machine learning models and the classic ARX model. In fact, the ARX model achieved the lowest average Root Mean Square Error (RMSE) across subjects in the test data when recursive method was used for multi-step prediction.

    UR - http://www.scopus.com/inward/record.url?scp=85051041853&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85051041853&partnerID=8YFLogxK

    M3 - Conference article

    VL - 2148

    SP - 97

    EP - 102

    JO - CEUR Workshop Proceedings

    JF - CEUR Workshop Proceedings

    SN - 1613-0073

    ER -