TY - GEN
T1 - Classification of Sepak Takraw Kicks Using Machine Learning
AU - Tan, Fu Yang
AU - Hassan, Mohd Hasnun Arif
AU - P. P. Abdul Majeed, Anwar
AU - Mohd Razman, Mohd Azraai
AU - Abdullah, Muhammad Amirul
N1 - Funding Information:
We thank Derrick Wilson and Paul Addison for advice on experimental procedures. The assistance of Mike Trolove and Gemma Garry in carrying out field work was most invaluable. Richard Watson and Nigel Bell reviewed this manuscript and suggested several valuable improvements. We thank Philip Froese, John Poul, and David Dodunski for the use of fields on their farms and for keeping us up to date with livestock grazing and the condition of the pastures used in our study. Some of the spider specimens were identified by Cor Vink (AgResearch, Lincoln). Catherine Cameron conducted the statistical analysis of the results. This work was funded by the Foundation for Research, Science and Technology. We also thank Mary Hiron for assisting in the final preparation of this manuscript for publication.
Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
PY - 2022
Y1 - 2022
N2 - Sepak Takraw has gained popularity over the years. Kinematics of the leg determine the quality and accuracy of the kick. However, the classification of Sepak Takraw kick using machine learning has never been explored. This study aims to classify the most typical kicks in Sepak Takraw namely the serve (or known as tekong), feeder and spike based on the leg’s kinematics using machine learning. Collegiate sepak takraw players participated in the data collection. The participants wore the inertial measurement unit sensor on their shank while performing the kicks. From the kinematics data recorded, several mathematical features were extracted and calculated. Machine learning algorithms such as the k-nearest neighbors (kNN), support vector machine (SVM), artificial neural networks (ANN), naive bayes (NB), random forest (RF), and logistic regression (LR) were applied to classify the types of kicks performed using fivefold cross-validation technique with 70% train data and 30% test data. It was found that ANN predicts all the test data correctly with 100% accuracy, followed by NB, SVM, RF and LR with 1 misclassification at 96.3% accuracy, kNN has the lowest prediction accuracy at 77.78%. This study shows that machine learning model is capable of classifying sepak takraw kicks. This can be used in training young athletes to ensure they perform the kicks properly, with correct skills.
AB - Sepak Takraw has gained popularity over the years. Kinematics of the leg determine the quality and accuracy of the kick. However, the classification of Sepak Takraw kick using machine learning has never been explored. This study aims to classify the most typical kicks in Sepak Takraw namely the serve (or known as tekong), feeder and spike based on the leg’s kinematics using machine learning. Collegiate sepak takraw players participated in the data collection. The participants wore the inertial measurement unit sensor on their shank while performing the kicks. From the kinematics data recorded, several mathematical features were extracted and calculated. Machine learning algorithms such as the k-nearest neighbors (kNN), support vector machine (SVM), artificial neural networks (ANN), naive bayes (NB), random forest (RF), and logistic regression (LR) were applied to classify the types of kicks performed using fivefold cross-validation technique with 70% train data and 30% test data. It was found that ANN predicts all the test data correctly with 100% accuracy, followed by NB, SVM, RF and LR with 1 misclassification at 96.3% accuracy, kNN has the lowest prediction accuracy at 77.78%. This study shows that machine learning model is capable of classifying sepak takraw kicks. This can be used in training young athletes to ensure they perform the kicks properly, with correct skills.
KW - Classification
KW - Leg’s kinematics
KW - Machine learning
KW - Sepak takraw kicks
UR - http://www.scopus.com/inward/record.url?scp=85116930141&partnerID=8YFLogxK
U2 - 10.1007/978-981-16-4115-2_26
DO - 10.1007/978-981-16-4115-2_26
M3 - Conference Proceeding
AN - SCOPUS:85116930141
SN - 9789811641145
T3 - Lecture Notes in Mechanical Engineering
SP - 321
EP - 331
BT - Human-Centered Technology for a Better Tomorrow - Proceedings of HUMENS 2021
A2 - Hassan, Mohd Hasnun
A2 - Ahmad (a) Manap, Zulkifli
A2 - Baharom, Mohamad Zairi
A2 - Johari, Nasrul Hadi
A2 - Jamaludin, Ummu Kulthum
A2 - Jalil, Muhammad Hilmi
A2 - Mat Sahat, Idris
A2 - Omar, Mohd Nadzeri
PB - Springer Science and Business Media Deutschland GmbH
T2 - Human Engineering Symposium, HUMENS 2021
Y2 - 22 February 2021 through 22 February 2021
ER -