Applying Time-Dependent Attributes to Represent Demand in Road Mass Transit Systems

Entropy

Teresa Cristóbal¹  Carmelo R. García¹  Gabino Padrón¹  Alexis Quesada-Arencibia¹  Javier Lorenzo-Navarro² 

[1] Institute for Cybernetics, Campus de Tafira, Las Palmas de Gran Canaria, University of Las Palmas de Gran Canaria, 35017 Las Palmas, Spain;University Institute of Intelligent Systems and Numeric Applications in Engineering, Campus de Tafira, Las Palmas de Gran Canaria, University of Las Palmas de Gran Canaria, 35017 Las Palmas, Spain;

关键词: clustering;    entropy;    attribute creation;    data mining;    intelligent transport systems;    mass transit systems;    demand;   

DOI  :  10.3390/e20020133

来源: DOAJ

【 摘 要 】

The development of efficient mass transit systems that provide quality of service is a major challenge for modern societies. To meet this challenge, it is essential to understand user demand. This article proposes using new time-dependent attributes to represent demand, attributes that differ from those that have traditionally been used in the design and planning of this type of transit system. Data mining was used to obtain these new attributes; they were created using clustering techniques, and their quality evaluated with the Shannon entropy function and with neural networks. The methodology was implemented on an intercity public transport company and the results demonstrate that the attributes obtained offer a more precise understanding of demand and enable predictions to be made with acceptable precision.

【 授权许可】

Unknown