Processing of the Time Series of Passenger Railway Transport in EU Countries

This article describes utilising the Eurostat railway passenger transport data as a time series in the Data Mining course lecturing at the university. The quarterly time series from 2004 to 2023 shows long-term increases or stable trends in passenger railway transport in European countries. The small decline after the economic crisis in 2008 is detected in the data. The highest decrease in passenger transport was in the second quarter of 2020 in all European countries caused by the COVID-19 pandemic. The number of transported passengers increased after the pandemic in 2020 and 2021 but is not fully at pre-COVID levels until the first quarter of 2023. Calculating the growth rate allows to compare the countries and annual changes. The practical example data shows that the decomposition of time series to trend, seasonal and residual parts must be processed separately for the part before the pandemic. The data mining software Orange helps create the processing and setting parameters quickly, like the number of values of sliding window for moving average to calculate the trend of time series. So students can concentrate on variants of processing and the correct interpretation of results. Orange was confirmed as an appropriate software for teaching the Data Mining course.