What is it about?
This article examines whether ski injuries can be predicted in order to be prevented. In order to do so, we used data mining and machine learning algorithms. First, we extracted features about skiers' behavior. Then we performed univariate and multivariate data analysis in order to identify risk factors. Finally, we performed CHAID decision tree and logistic regression methods which provided us predictive model which is interpreted.
Featured Image
Why is it important?
This article is important because it 1) extracts features or characteristics of skiers from RFID data, 2) utilize data mining and machine learning algorithms in order to predict ski injuries, 3) identify risk factors and interpret them. One of the main findings of this article is that we recognized ski injuries as early-failure events. Additionally, one can find univariate and multivariate analysis of extracted features, and interpretation of that analysis.
Perspectives
I hope this article will help people understand that ski injuries can be prevented just by analyzing skiing behavior. This article provide insight from Mt. Kopaonik, i.e. what behavior is considered as risky behavior. Although results can not be generalized, methodology can be applied to different mountains. As a major limitation of this article we state lack of width in data (weather and demographic data were not available).
Sandro Radovanovic
University of Belgrade
This paper makes useful suggestions on how to stay safe in ski resorts. It also provides a methodology how to work with highly imbalanced datasets, i.e. how to segment objects into more heterogeneous decision nodes.
Boris Delibasic
University of Belgrade - Faculty of Organizational Sciences
Read the Original
This page is a summary of: Ski injury predictive analytics from massive ski lift transportation data, Proceedings of the Institution of Mechanical Engineers Part P Journal of Sports Engineering and Technology, September 2017, SAGE Publications,
DOI: 10.1177/1754337117728600.
You can read the full text:
Resources
Contributors
The following have contributed to this page
