Search

Public transport ridership has fallen drastically over the past few years, resulting in an increased number of private vehicles on the road, negatively impacting the road infrastructure with escalated levels of traffic congestion, pollution, road accidents, etc. Policy interventions provide a viable solution to the problem of ridership as they influence the decision-making of the public. In this article, an attempt was made to model the choice behaviour of employees working in the city of Thiruvananthapuram, India based on soft computing techniques. The choice data were collected through a revealed preference survey and mode choice models were developed using Artificial Neural Networks (ANN), Neuro-Fuzzy, and Fuzzy logic taking socio-economic characteristics, trip characteristics, and the identified latent factors as inputs. The results from the developed models are encouraging and they show good statistical validity. To formulate passenger attraction policies, the major latent factors that affect public bus ridership rates were examined. Consequently, suitable transport improvement policies are recommended based on the most influencing latent factors. Finally, based on the results of the sensitivity analysis, several policy interventions to enhance public transport ridership are recommended. [ABSTRACT FROM AUTHOR]