INDIVIDUAL RESPONSES TO RUNNING SHOES: AN INVESTIGATION USING UNSUPERVISED MACHINE LEARNING

Authors

Djordje Slijepcevic, Institute of Creative Media Technologies, Department of Media & Digital Technologies, St. Pölten University of Applied Sciences, St. Pölten, AustriaFollow
Fabian Horst, Department of Training and Movement Science, Institute of Sport Science, Johannes Gutenberg-University Mainz, Mainz, GermanyFollow
Matthias Zeppelzauer, Institute of Creative Media Technologies, Department of Media & Digital Technologies, St. Pölten University of Applied Sciences, St. Pölten, AustriaFollow
Wolfgang Immanuel Schöllhorn, Department of Training and Movement Science, Institute of Sport Science, Johannes Gutenberg-University Mainz, Mainz, GermanyFollow

Category

Running & Jumping

Document Type

Paper

Abstract

This study aimed to investigate the movement patterns of individual subjects during running across three (standardized) shod and one barefoot condition, employing t‑distributed stochastic neighbor embedding (t-SNE) and k-means cluster analysis. The t‑SNE analysis revealed three individual response strategies to different footwear conditions: (1) maintaining consistent ground reaction force (GRF) patterns across all conditions; (2) maintaining consistent GRF patters across two or three conditions; and (3) exhibiting changes in GRF patterns across all conditions. The cluster analysis showed that subjects were more often grouped into the same cluster when footwear conditions were more similar. The results indicate that only one group of two and one group of three subjects (out of 30 subjects) exhibited similar GRF patterns across all footwear conditions.

Recommended Citation

Slijepcevic, Djordje; Horst, Fabian; Zeppelzauer, Matthias; and Schöllhorn, Wolfgang Immanuel (2024) "INDIVIDUAL RESPONSES TO RUNNING SHOES: AN INVESTIGATION USING UNSUPERVISED MACHINE LEARNING," ISBS Proceedings Archive: Vol. 42: Iss. 1, Article 247.
Available at: https://commons.nmu.edu/isbs/vol42/iss1/247