A machine learning approach to study the effect of parking pricing on transport mode choice to work using 2022 NHTS dataset

The amount of traffic on city streets is growing, which results in traffic congestion, lowers road safety, and makes it more difficult to find a parking space. According to the 2022 National Household Travel Survey (NHTS) – California, over 80% of commuters drove to work using single-occupant vehicles (SOV) which causes several issues including decreased urban liveability, increased traffic congestion, parking spaces in CBD, and GHG emissions. Past studies considered employer-paid parking, private parking space sharing, cruising, curbside parking, cargo/heavy vehicle parking spaces, and vacant parking spaces with detection devices to reduce the use of SOV, traffic congestion in urban areas, and promote sustainable transportation using conventional techniques. However, limited studies have been conducted on the effect of parking pricing on travel mode choice (TMC) using modern techniques. Therefore, the current study aims to study the effect of parking pricing on work-related TMC using conventional and modern techniques. For the statistical analysis, some of the variables related to the scope of the study from the 2022 NHTS dataset were used. Several machine learning (ML) algorithms such as random forest (RF), support vector machine (SVM), decision tree (DT), and Neural Network (NN) were applied to study the relation. The outcomes show that modern techniques outperformed traditional techniques. Among all ML algorithms, RF outperformed with an accuracy of 97.5% and precision of 97.6% followed by DT with 91.7% accuracy and 91.8% precision. Feature importance was applied to study the most significant feature and claims that PARK2 (paid for the parking on the trip) was the most significant feature stating that the payment at the workplace significantly influenced the TMC to work. Urban planners and policymakers need to develop their policies based on the study outcomes (importance of ML algorithms for TMC predictions and imposing parking fees at the workplace) that can promote sustainable transportation systems. Built environment variables, fear of safety, and insecurity should be considered by future researchers to predict TMC and promote sustainable transportation systems.