Page 69 - THEi Student Applied Research Presentations 2024
P. 69

THE ESTIMATION OF RUNNING ECONOMY AND
    MOTION ANALYSIS OF RUNNERS USING MACHINE

    LEARNING ON MOTIONMETRIX DATA


    ABSTRACT               Poster


    RESEARCH BACKGROUND

    This study investigates whether there is a particular pattern among runners in achieving
    efficient running economy and running performance. Could there be a relation between
    the running biomechanisms and what would be the potential intrinsic factors affecting
    the running economy?

    METHODOLOGY                                                                  THAPA CELCIA
    A total of 59 athletes, aged 21 to 53 years, participated in this study conducted at
    RunLap Research Centre. Subjects were asked to perform the following tasks: (1) Foot   BSocSc (Hons) in Sports and
    posture index, (2) Ankle dorsiflexion range of motion (ROM), (3) Reactive strength   Recreation Management
    index (RSI), (4) Single-leg squat, (5) Hip abduction strength, and (6) Running gait   Department of Sport and Recreation
    assessment (MotionMetrix).

    FINDINGS

    Out of the three machine learning models (XGB, SVM, ANN), the study demonstrated
    that XGB is the most accurate for predicting running performance. Moreover, the   OBJECTIVES
    calculated cross-validated mean absolute error (MAE) of 3.757 and MAE of 4.371 fits
    under the 10% potential error acceptance range in predicting the running duration. The   To use MotionMetrix to analyse the
    most significant features of importance were training pace, weekly distance, right knee   running  motion  through  kinematics,
    alignment, and overstride.                                                   such as joint loading, gait symmetry,
                                                                                 and  stride  length.  Furthermore,  with
    ABOUT THE INVESTIGATOR                                                       the data collected, this research aims
                                                                                 to  identify  the  influential  factors  or
    I’m THAPA Celcia. As an active sportsperson, I enjoy the thrill and all the other emotions   patterns  among runners  on optimal
    sports give me. Hence, I genuinely like helping people, whether at the clinic or during   running performance and the economy
    on-field support. Moreover, when working with sports teams, the atmosphere is one   of running.
    of the reasons that drives me to push myself further as a sports therapist. To my
    supervisor, Mr Indy HO, thank you for broadening my horizons by introducing machine
    learning to me through this research project.

































 Student Applied Research Presentations 2024                                  Student Applied Research Presentations 2024  69
   64   65   66   67   68   69   70   71   72   73   74