Analysing the Unseen: Leveraging Data Analytics to Combat the Societal Challenge of Doping in Sports

Sports officials worldwide face societal challenges due to the unfair nature of fraudulent practices performed by dishonest athletes. One such practice is sample swapping, where athletes exchange their doped sample with a clean one to avoid testing positive. The current detection method for such cases involves biochemical testing, i.e., DNA typing, which is expensive and time-consuming, making it difficult for anti-doping organisations to operate within their budgets. Therefore, there is a need to explore alternative methods to improve decision-making. To address this issue, we propose a Subsampling-based Convolutional Neural Network (SCNN) that enables automatic feature learning while preserving the longitudinal aspect of the athlete profile. Our model can identify swapped samples and outperforms the current state-of-the-art method and baseline models in the anti-doping community. We have evaluated our model on real-world sample swapping cases, and the results are promising, contributing to data analytics development in the context of anti-doping analysis.