PREDICTION OF ATHLETIC ABILITY AND TALENT DETECTION OF RURAL AND TRIBAL CHILDREN USING THE K-MEAN CLUSTERING ALGORITHM

Background - Talent Identification and Development (TID) in sports is a continuum and multidimensional process of identifying, selecting, and developing potentialathletes (at an early age) in specific sports based on their present performance or abilities. Government-bagged schemes are proven effect...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:NeuroQuantology 2022-01, Vol.20 (1), p.1094
Hauptverfasser: Chaware, Utsav, Das, Anindita
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Background - Talent Identification and Development (TID) in sports is a continuum and multidimensional process of identifying, selecting, and developing potentialathletes (at an early age) in specific sports based on their present performance or abilities. Government-bagged schemes are proven effective in the TI process but studies claim the rural and tribal athletes fail to claim due to the operational limitations. Therefore Talent Detection (TD) plays a vital role, however, research suggests with limited experts, coaches, and selectors and a lack of robust mechanism the selection is not data-driven and could be biased. Henceforth, the use of machine learning could be the solution to detect talent. Aim- The study aimed to develop a screening and classification model based on the anthropometric and physical parameters of ruraland tribal children using K-means clustering, an unsupervised clustering method. Material and Methods- 240 tribal and rural children aged between 15-17 years from tribal districts of Madhya Pradesh with no formal training were assessed for demographic, anthropometric and physical variables based on the ‘Khelo India’physical fitness and talent identification scheme. Data was analysed and clustering was done using a K-mean unsupervised algorithm in Python. Result and Discussion- Results of PCA show 4 physical attributes explaining a total variability of 70.18%. Athletes were classified into 3 clusters; sports fit(27.7%), athletic in at least one ability(23.7%), and non-atheltic(46.6 %). Further, ANOVA between the groups reveals significant differences (p0.05). However, the results were apparent between non-athletic and other groups. Conclusion- The present study provides a data-bagged classification system that clusters the low performers (non-athletic group) from another group, unloading the burden of mass scrutiny and data analysis, particularly for rural and tribal children as most of them have no formal skill training in any sports.
ISSN:1303-5150
DOI:10.48047/NQ.2022.20.1.NQ22393