Predicting motorcycle customization preferences using machine learning
Recent developments in artificial intelligence have expanded the role of machine learning in analyzing human behavior across various domains. Despite this progress, few studies apply supervised learning to behavioral prediction in vehicle personalization. This study explores potential by examining i...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | en |
| Published: |
IEEE
2025
|
| Subjects: | |
| Online Access: | https://umpir.ump.edu.my/id/eprint/47003/1/Predicting_Motorcycle_Customization_Preferences_using_Machine_Learning-2.pdf https://umpir.ump.edu.my/id/eprint/47003/ https://doi.org/10.1109/COMNETSAT68601.2025.11324650 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Recent developments in artificial intelligence have expanded the role of machine learning in analyzing human behavior across various domains. Despite this progress, few studies apply supervised learning to behavioral prediction in vehicle personalization. This study explores potential by examining individual tendencies in choosing between modified and factory-original motorcycle. A dataset comprising 292 respondents was compiled, capturing variables such as age, social environment, financial capacity, and exposure to automotive communities and content. The classification model was developed using the Random Forest algorithm, Support Vector Machine and Logistic Regression with 5-fold Cross validation. Random forest was chosen in this research as a primary algorithm because it is superior and its efficiency in processing heterogeneous data. The resulting model demonstrates Random Forest predictive capabilities with an accuracy of 82.54% outperforming both their alternative methods, supported by balanced precision, recall, and F1-score metrics. Further analysis revealed that user interest in automotive topics, influence from peers, budget allocation, community engagement, and digital media usage were among the most influential factors. These findings highlight the complexity of consumer behavior in the automotive sector and suggest that machine learning can offer valuable insights for stakeholders seeking to tailor services and strategies based on user preferences. |
|---|