Publication: A classification model on types of cycling based on cycling behavior
Loading...
Date
2023-07-14
Authors
Chieng, Hui Zhi
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Cycling has gained popularity as a form of exercise, and individuals use various cycling techniques and habits depending on their interests. On the basis of cycling behavior, there has been little study on classification models for different types of cycling. Therefore, the goal of this project is to create a supervised learning approach to categorize different cycling kinds using data on cycling performance behavior. In order to categorize different types of cycling based on cycling behavior, this study compares the performance of several classification algorithms at 10-fold cross validation mode, including BayesNet, SMO, IBk, KStar, RandomizableFilteredClassifier, and RandomTree, on two case studies taken from publicly available data. For CS1 and CS2, respectively, the baseline classifier, ZeroR, is 51.52% and 38.64%. CS1 was typically accurately identified using a variety of methods, including BayesNet, SMO, IBk, and RandomTree. IBk and KStar, with a maximum accuracy of 71.59%, are the algorithms that demonstrate the highest accuracy for CS2. The results help us understand cycling activities better, making it easier for cyclists to develop individualized training plans and make decisions in a variety of settings