Predicting Cycling Performance Using Machine Learning
Electronic Theses and Dissertations
Item Files
Item Details
- abstract
- For over two decades, professional cyclists have been relying on power meters in their training. These devices provide an accurate and reliable measurement of the effort made while cycling. Such devices, however, are costly for the majority of cyclists and they often come with an overwhelming variety of features, which can be confusing to non-professionals. Several popular online platforms use physics-based equations for estimation of cycling power and they rely on several assumptions about environment. At the same time, these platforms have access to large collections of workout data along with the personal characteristics of riders, which could be used for the data-driven predictions of cycling power.
- subject
- cycling
- machine learning
- performance
- power
- prediction
- sports
- contributor
- Khuri, Natalia (committee chair)
- Cañas, Daniel (committee member)
- Pauca, Paúl (committee member)
- date
- 2020-05-29T08:35:52Z (accessioned)
- 2021-05-28T08:30:12Z (available)
- 2020 (issued)
- degree
- Computer Science (discipline)
- embargo
- 2021-05-28 (terms)
- identifier
- http://hdl.handle.net/10339/96810 (uri)
- language
- en (iso)
- publisher
- Wake Forest University
- title
- Predicting Cycling Performance Using Machine Learning
- type
- Thesis