Train Stations and Stops Optimal Localization at Rail Lines Regional Significance
Abstract: The paper deals with the search for a system tool for the station and stops localization at rail lines regional significance based on data from the Census 2011 (SLDB), Territorial Analytical Data of the Czech Republic 2014 (ÚAP) and additional information especially from maps. Europe has been dealing with the issue of rail lines regional significance for decades, and each country has approached the question of further existence or non-existence in other ways. In the Czech Republic, this issue has been particularly emphasized at the last twenty years, so in the time this type of transport infrastructure and trains in operation has been in the end of a technical lifetime, and it was necessary to decide what further. The database of municipalities has been tested by correlation, clustering and regression analysis, with the aim of maximizing the number of passengers at the particular station or stop, thus maximizing the overall potential of the rail line for possible reconstruction. In the database, cluster analysis methods estimate static mixture components, DBSCAN and K-means are tested to create clusters of character-like municipalities. The most successful algorithm that determined the resulting shape of clusters is the K-means. The main output of the paper is a summary of recommendations with a decision tree and an equation to maximize the passengers use of a particular train station or stop including recommendations for future data collection.
