USP Electronic Research Repository

Comparative Analysis of Classification Techniques on Olympic Games Datasets

Nand, Ravneil and Chand, Ashneel and Reddy, Emmenual (2023) Comparative Analysis of Classification Techniques on Olympic Games Datasets. [Conference Proceedings]

Full text not available from this repository.


Olympic game is a prestigious ceremony that occurs after every four years. However, due to the spread of coronavirus in 2020, the game was held in 2021, which is post-Covid. The main aim of this research is to find out if there was a difference in the performance of nations in Rio 2016 Olympics (pre-Covid) and Tokyo 2020 Olympics (post-Covid). Statistical analysis is carried out to find the correlation between the different variables. One of the highly correlated variables (Gold Tally) is removed while performing the classification analysis. The idea is to see if the classifiers are able to do the comparative analysis without it or not. The classification algorithms utilized in this research are Decision Table, Decision Tree, Naïve Bayes, and Random Forest. The datasets used in this research are imbalanced sets, which were later transformed to balance sets through under-sampling. Random Forest was able to give 100 % accuracy in both datasets whereas the True Positive Rate (TPR) was also 100%. After doing the comparative analysis it was found that irrespective of pre and post-Covid, the performance of athletes did not change. This paves the way for other researchers to investigate if Covid had any impact on the performance of the athletes or not. In the future, more vast variables will be investigated to do a more detailed comparative analysis.

Item Type: Conference Proceedings
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Centre for Flexible Learning (CFL)
School of Information Technology, Engineering, Mathematics and Physics (STEMP)
Depositing User: Emmenual Reddy
Date Deposited: 31 Jan 2024 23:55
Last Modified: 06 Mar 2024 21:48

Actions (login required)

View Item View Item