Classification of high performance archers by means of bio-physiological performance variables via k-nearest neighbour classification model

The present study classified and predicted high and low potential archers from a set of bio-physiological variables trained via a machine learning technique namely k-Nearest Neighbour (k-NN). 50 youth archers drawn from various archery programmes completed a one end archery shooting score test. Bio-physiological measurements of systolic blood pressure, diastolic blood pressure, resting respiratory rate, resting heart rate and dietary intake were taken. Multiherachical agglomerative cluster analysis was used to cluster the archers based on the variables tested into low, medium and high potential archers. Three different k-NN models namely fine, medium and coarse were trained based on the measured variables. The five-fold cross-validation technique was utilised in the present investigation. It was shown from the present study, that the utilisation of k-NN is non-trivial in the classification of the performance of the archers