A COMPARATIVE ANALYSIS OF THE FACTORS INFLUENCING UNIVERSITY STUDENTS' MICROMOBILITY PREFERENCES USING K-NEAREST NEIGHBORS AND LOGISTIC REGRESSION MODELS

Shared micro-mobility services have swiftly become widely adopted in major urban centers globally. In particular, individuals are encouraged to transition to environmentally friendly modes of transportation to support a sustainable transportation system. For this reason, the tendencies and potential of individuals to use micro-mobility vehicles are being investigated. This paper focused on university students, analyzing their preferences for using micromobility vehicles, particularly for first-mile or last-mile trips in terms of gender and travel time variables. In the study, k-Nearest Neighbors (kNN) and Logistic Regression (LR) algorithms are used in machine learning approach and they were compared. A face-to-face survey was conducted with 150 students randomly to measure the potential use of micromobility vehicles among university students. As a result, LR model is better than kNN model according to the accuracy of the models, 0,63 and 0,43 respectively. On the other hand, 51,82% of male students and 62,50% of female students participating in our study reported that they are not inclined to prefer micromobility vehicles at any stage of their trips, and the main challenge for the potential users is safety.