Bus service quality prediction and attribute ranking using probabilistic neural network and adaptive neuro fuzzy inference system

Abstract

This study applies probabilistic neural network (PNN) and adaptive neuro fuzzy inference system (ANFIS) to develop bus service quality (SQ) prediction model based on the preferences stated by users (on a scale of 1 to 5). A questionnaire survey is conducted and a data set from the survey is prepared to develop the SQ prediction model using PNN and ANFIS. Results show that ANFIS produced better prediction than PNN. The research is further extended to include ranking of the SQ attributes according to their impact on the overall result from the developed model. Attributes such as punctuality and reliability, seat availability, and service frequency were found to be the top three attributes that mostly affect the decision making process of the users. This study can aid service providers in improving the most important attributes of bus service to develop the quality of service, thereby increasing transit ridership.

A Review of Methodological Approaches and Modeling Techniques in Service Quality Evaluation of Surface Transportation during the Last Decade

“…The different types of neural networks employed in SQ studies and often compared to each other or other techniques include ANN [37], comparison between ANN and DT [36] comparison between ANFIS and PNN [24], comparison between PNN, PRNN, and GRNN [25] and ANFIS [22].…”

The Application of Adaptive Neuro-Fuzzy Inference System and Fuzzy Delphi Technique to Assess Socio-Economic Impacts of Construction of Rural Roads

“…Membership functions of the subcriteria "involvement in social-gathering outside the village" (Q SGOV ) of quality of neighbourhood criteria are depicted as below in Figure 6. The parameters employed for ANFIS models are obtained from the literature (Islam et al, 2016;Keshavarzi et al, 2017). Table 7.…”

“…It is widely used in condition identification (Hosseinlou and Sohrabi, 2009), decision making (Pamučar et al, 2013), prediction modeling (Lee et al, 2015), etc. It provides results with a tolerance of ambiguity, uncertainty, approximation and handle complex social and human systems comprehensively by utilizing linguistic information in the form of human perception and measured data (Islam et al, 2016), as well as it is time and cost effective. Thus, it is well understood that the ANFIS technique can overcome the existing research shortcomings in the available techniques employed to assess socio-economic impacts thoroughly.…”

Development of neuro‐fuzzy‐based multimodal mode choice model for commuter in Delhi

“…Later, researchers have incorporated AI techniques such as ANN [17][18][19], FL [20][21][22], NF [23][24][25], support vector machine and random forest decision tree [26] to develop mode-choice models. The use of AI techniques in travel demand modelling began in 1960.…”

“…This paper showed that NF approach has a great potential of successful application to a wide range of transportation problems. Islam et al [18] developed a bus service quality prediction model using ANFIS and probabilistic NN. ANFIS model predicted the outcome more accurately with commuters being most sensitive to attributes such as punctuality and reliability, seat availability and service frequency.…”