Evolutionary k-means Graph Clustering Method to Obtain the hub&spoke Structure for Warsaw Communication System
Jarosław Tomasz Stańczak, Jan W. Owsiński, Barbara Mażbic-Kulma, Aleksy Barski, Krzysztof Sęp · Annals of Computer Science and Information Systems · 2023
The k-means method is one of the most frequently used clustering methods due to its efficiency and ease of modification and adaptation to the problem being solved.This paper presents modification of k-means method used for clustering in graphs.The method is presented on the example of generating the hub&spoke structure in the graph of public transport connections in Warsaw.Optimization of the public transport is one of the most important tasks for large cities.An efficient transport system is very important for its inhabitants.One of possible solutions is introducing the idea of hub&spoke to transport system.In this approach it is important to detect main stations, called hubs, which will create axes of high-speed connections (city trains, metro, high-speed trams), from which passengers can transfer to slower local connections to get to their rather close destinations.In the presented approach we propose to find locations of such main changeover stations using an evolutionary k-means algorithm.