Running-Related Injury Classification For Professional Runners
Running being a form of healthy physical activity which is prone to injuries if performed excessively or with incorrect posture. Previous studies have considered risk factors of running-related injuries (RRI) to be limited and have multifactorial origins. However, little is discussed on prediction...
Saved in:
Main Author: | |
---|---|
Format: | Monograph |
Language: | English |
Published: |
Universiti Sains Malaysia
2021
|
Subjects: | |
Online Access: | http://eprints.usm.my/55781/1/Running-Related%20Injury%20Classification%20For%20Professional%20Runners.pdf http://eprints.usm.my/55781/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my.usm.eprints.55781 |
---|---|
record_format |
eprints |
spelling |
my.usm.eprints.55781 http://eprints.usm.my/55781/ Running-Related Injury Classification For Professional Runners Lingam, Darwineswaran Raja T Technology Running being a form of healthy physical activity which is prone to injuries if performed excessively or with incorrect posture. Previous studies have considered risk factors of running-related injuries (RRI) to be limited and have multifactorial origins. However, little is discussed on prediction parameters to be considered when studying the type of potential injury risks that may affect a particular runner. This study aims to investigate the qualities of RRI dataset for reliable running-injury classification analysis with WEKA and also to establish an appropriate classification model for RRI in professional runners. The data from 74 professional runners were collected from Kaggle repository. This dataset consisted of injured and uninjured classes measured by data attributes (nr. rest days, total km Z3-Z4-Z5-T1-T2, total km Z3-4, total km Z5-T1-T2, total hours alternative training, nr. strength trainings, avg exertion, avg training success, and avg recovery). Classification analyses were performed on study data using BayesNet, RandomForest, J48, RandomTree, REPTree, and IBk algorithms in WEKA toolkit. The RRI dataset was pre-processed to filter outliers and extreme values as well as irrelevant data attributes prior to the classification. Findings revealed that three best classifier algorithms with the highest accuracies to classify runners into the category of uninjured and injured are BayesNet (98.6457%), RandomForest (98.0107%), and (unpruned) J48 (97.1002%). This research is a step forward in predicting a probable RRI in professional runners using a data mining approach. Universiti Sains Malaysia 2021-07 Monograph NonPeerReviewed application/pdf en http://eprints.usm.my/55781/1/Running-Related%20Injury%20Classification%20For%20Professional%20Runners.pdf Lingam, Darwineswaran Raja (2021) Running-Related Injury Classification For Professional Runners. Project Report. Universiti Sains Malaysia, Pusat Pengajian Kejuruteraan Mekanik. (Submitted) |
institution |
Universiti Sains Malaysia |
building |
Hamzah Sendut Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Sains Malaysia |
content_source |
USM Institutional Repository |
url_provider |
http://eprints.usm.my/ |
language |
English |
topic |
T Technology |
spellingShingle |
T Technology Lingam, Darwineswaran Raja Running-Related Injury Classification For Professional Runners |
description |
Running being a form of healthy physical activity which is prone to injuries if performed excessively or with incorrect posture. Previous studies have considered risk factors of running-related injuries (RRI) to be limited and have multifactorial origins.
However, little is discussed on prediction parameters to be considered when studying the type of potential injury risks that may affect a particular runner. This study aims to investigate the qualities of RRI dataset for reliable running-injury classification analysis with WEKA and also to establish an appropriate classification model for RRI in professional runners.
The data from 74 professional runners were collected from Kaggle repository. This dataset consisted of injured and uninjured classes measured by data attributes (nr. rest days, total km Z3-Z4-Z5-T1-T2, total km Z3-4, total km Z5-T1-T2, total hours alternative training, nr. strength trainings, avg exertion, avg training success, and avg recovery). Classification analyses were performed on study data using BayesNet, RandomForest, J48, RandomTree, REPTree, and IBk algorithms in WEKA toolkit. The RRI dataset was pre-processed to filter outliers and extreme values as well as irrelevant data attributes prior to the classification. Findings revealed that three best classifier algorithms with the highest accuracies to classify runners into the category of uninjured and injured are BayesNet (98.6457%), RandomForest (98.0107%), and (unpruned) J48 (97.1002%). This research is a step forward in predicting a probable RRI in professional runners using a data mining approach. |
format |
Monograph |
author |
Lingam, Darwineswaran Raja |
author_facet |
Lingam, Darwineswaran Raja |
author_sort |
Lingam, Darwineswaran Raja |
title |
Running-Related Injury Classification For Professional Runners |
title_short |
Running-Related Injury Classification For Professional Runners |
title_full |
Running-Related Injury Classification For Professional Runners |
title_fullStr |
Running-Related Injury Classification For Professional Runners |
title_full_unstemmed |
Running-Related Injury Classification For Professional Runners |
title_sort |
running-related injury classification for professional runners |
publisher |
Universiti Sains Malaysia |
publishDate |
2021 |
url |
http://eprints.usm.my/55781/1/Running-Related%20Injury%20Classification%20For%20Professional%20Runners.pdf http://eprints.usm.my/55781/ |
_version_ |
1751537274703577088 |
score |
13.18916 |