Forecasting of Moscow metro passenger’s traffic applying artificial neural networks
Seminars
5 November 2015, 15:00, LIT room 310
1. V.V.Ivanov, E.S.Osetrov "Forecasting of Moscow metro passenger’s traffic applying artificial neural networks"
Abstract
In the work we developed methods for the prediction of passenger’s traffic in Moscow metro using artificial neural networks. To this end, the analysis and the selection of those factors that have a major impact on passenger traffic in the subway have been realized. For data that correspond to the daily volumes of passenger’s traffic in weekdays, demonstrated the fundamental possibility of short-term forecasting with acceptable accuracy.
2. V.V.Ivanov, E.S.Osetrov "Forecasting of Moscow metro passenger’s traffic based on neural networks with prior filtering parsed data"
Аннотация
In the paper [1] we have developed an approach for forecasting of passenger traffic in the Moscow metro by using artificial neural networks. For data that correspond to the daily volumes of passenger’s traffic in weekdays, it has been demonstrated principal possibility of short-term forecasting with acceptable accuracy. In the present work it is shown that the use of wavelet filtering for realized values of daily traffic can significantly improve the accuracy of prediction and increase the time horizon.