Multilayer Perceptron Modelling of Travelers Towards Park-and-Ride Service in Karachi

The imbalance between public and private transport causes congestion. Currently, congestion is due to individuals driving their automobiles to work in Karachi central business districts (CBDs). Therefore, the park and ride (P&R) service has been utilized widely in several countries as part of travel demand management (TDM). Consequently, P&R has proved successfully reduced congestion and difficulties to locate parking spots in the urban center. Travelers cannot be persuaded to adopt P&R without knowing their travel pattern. Accordingly, a travel behavioral survey was conducted, to eliminate imbalances between public and private mobility. Therefore, modal choice models were to identify the variables influencing the decision to accept P&R service of single-occupant vehicles (SOV). Data were collected by an adapted self-administered questionnaire through a survey approach. Mode choice models developed through multilayer perceptron (MLP) of artificial neural network (ANN) approach by using statistical package for the social sciences (SPSS) version 22. The research findings were more towards the socio-demographic factors. Furthermore, travel time, travel expenses, environmental protection, avoid mental stress, parking problem, vehicle sharing, and travel directly from home to office were found significant variables. In conclusion, the SOV users may be encouraged to move into P&R services by overcoming these influencing elements and balance push and pull measures of TDM. Thus, policymakers can benefit from study results and provide a base for future studies on sustainable modes of public transportation. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.