Abstract
This article presents an approach to optimizing existing transport routes in Almaty based on data reflecting changes in population density throughout the day. Aggregated data from a telecommunications operator, geographic information from OpenStreetMap, and transport data from CityBus were utilized to identify areas of high population activity. The application of clustering methods (DBSCAN) and machine learning algorithms (LightGBM) enabled the evaluation of route efficiency and the identification of the primary influencing factors. The analysis revealed that the number of stops is the most significant factor, and optimizing their locations can substantially enhance the efficiency of the transport system. The findings underscore the potential of the proposed approach to improve urban transport systems and align them more closely with the real needs of the population.