University of Twente Student Theses


Investigation of Quality Measures in Cyclists’ Dataset Using Dimensionality Reduction Techniques

Solovyeva, Olga (2022) Investigation of Quality Measures in Cyclists’ Dataset Using Dimensionality Reduction Techniques.

[img] PDF
[img] HTML
Abstract:Multiple external and internal factors could influence the performance of cyclists: heartbeat, age, gender, speed, elevation, wind speed, temperature, and distance, among others. Those factors are essential for planning on how to enhance the overall athlete’s performance. However, certain factors could influence the performance more heavily than others. To gain insights into how those factors could intertwine, multidimensional visualization techniques could be useful when exploring visual patterns. In particular, dimensionality reduction techniques may uncover more details on why some athletes perform at a high level, whilst others struggle. With an enormous number of existing dimensionality reduction techniques, this research proposes to find the most qualitative technique in distinct datasets, including a cyclists’ dataset. The results show that t-SNE shows outstanding performance in terms of neighborhood and distance preservation and has the potential to be used with clustering algorithms to demonstrate new insights into the cyclists’ data. Since dimensionality reduction techniques for cycling data are not well explored by scientific literature, this opens an opportunity for research in this field that could add substantial contributions to those who would be interested in improving cycling behavior.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page