<![CDATA[Marathons are becoming an increasingly popular form of exercise and social interaction. Participants who choose race categories based on the mileage provided, such as 6K, 7.9K, and 11K, according to personal preference. However, this category selection has not been analyzed based on participant characteristics, even though this information is important for organizers to support promotional strategies, and segmentation of participants. This study aims to predict marathon category selection based on demographic characteristics, namely age and gender, by applying Decision Tree and Random Forest machine learning algorithms. The dataset used is primary data from two events, namely RSDK Berlari with a total of 1091 data and Skybridgefunrun with 1519 data. The results show that the Decision Tree algorithm gets an accuracy of 56.81%, and the Random Forest algorithm is 57.38%. With these results, it shows that the Random Forest algorithm has higher accuracy than the Decision Tree algorithm, with accuracy reaching 57.38%. However, the model tends to be biased towards the 7.9K category, with recall reaching 94%, while the 6K and 11K categories are very low. Then, feature importance analysis shows that the most influential factor on category selection is age, while gender is smaller. This research provides insight for event organizers in designing promotional strategies and participant segmentation more precisely.]]>
Copyrights © 2025
