Comparison of Parametric and Non-Parametric Methods for Modeling Establishment-Level Freight Generation

Abstract

The traditional parametric modeling approaches used to predict freight generation (FG), such as ordinary least squares (OLS), suffer from limitations because of the realistic possibility of violating basic assumptions such as linearity or data distribution. This problem is multidimensional because of the need to model numerous industry sectors in the freight system and the possibility of using different explanatory variables; little guidance is currently available on which modeling methodology is suitable for a particular case. The non-parametric models that could solve several limitations with traditional FG models are rarely examined in their predictive ability. This paper offers insights into this frequent research question by comparing the performance of various modeling methodologies that can be used for predicting FG: OLS, weighted least squares, robust regression (RR), seemingly unrelated regression (SUR), multiple classification analysis (MCA), and support vector regression (SVR). To this effect, the research carried out in this study uses a freight dataset of 432 establishments across Kerala State, India. The model outputs are validated using resubstitution and cross-validation methods, and the prediction errors are quantified using root-mean-square error and mean absolute error. The validation results show that the non-parametric SVR models are better alternatives in developing state-, regional- and industrial segment-level models. The MCA models are more precise in predicting FG for suburban models. RR models provide a better predictive ability for modeling FG in some industrial segments. Overall comparison and result interpretations suggest that the non-parametric models are superior in relation to predicting FG. At the same time, RR seems to be the only parametric modeling approach that can provide comparable model performance to non-parametric models.

Examining the Effects of Local Sample Sizes on Spatial Transferability of Freight Production Models

“…MCA is a frequently employed straightforward method that presents the average FP rates in tabular format and does not follow any functional form ( 21 , 22 , 40 , 41 ). This technique is comparable to OLS, but dummy variables are employed.…”

“…In addition, they estimated FG rates using multiple classification analysis (MCA) and developed nomographs for different industrial classes. Using parametric and non-parametric modeling techniques, Balla et al ( 22 ) developed FG models for various industrial classes located in the same Indian cities. This study classified the industries using an a posteriori classification scheme (not ISIC classification like the previous study).…”

Developing B2C freight trip generation model from Courier Express Parcel establishments: A case study of two Indian cities

“…Apart from that, the areas considered in the past studies include Belgium (Beckers et al 2022), Brazil (Oliveira et al,2017), Italy (Comi and Nuzzolo, 2016), Netherlands (Iding et al, 2002), Portugal (Alho & de Abreu e Silva, 2015; Alho & de Abreu e Silva, 2017), Sweden (Sanchez-Diaz et al, 2013;Sánchez-Díaz, 2017) and USA (Bastida & Holguín-Veras, 2009;Holguín-Veras et al 2011;Jaller et al, 2015).FG/FTG studies are in their infancy both in India and other developing nations, although an increase in research for examining the FG and FTG has been witnessed in India in the last few years. Freight tour activities, crop production and attraction, freight production and attraction from establishments were analysed and modelled in those studies (Balla et al, 2023;Dhulipala & Patil, 2020;Pani and Sahu, 2018;Pani and Sahu, 2020;Venkadavarahan & Marisamynathan, 2021;Venkadavarahan et al, 2020;Venkadavarahan & Marisamynathan, 2022). Few studies investigated B2C freight in Belgium, Rome and the USA (Beckers et al, 2022;Comi and Nuzzolo, 2016;Rodrigue, 2022).…”

“…found non-parametric (MCA and Support Vector Regression) models better for FG predictions than parametric ones (OLS, Weighted Least Square, Robust Regression, Seemingly Unrelated Regression). RMSE and MAPE values were used as performance parameters for comparison purposes (Balla et al, 2023). Venkadavarahan and Marisamynathan developed FTG models using MCA & studied the temporal pattern of daily freight trips from/to establishments of Trichy.…”

Characterising and Modelling Urban Freight in Developing Economies

“…The techniques used to model FG/FTG include ordinary least squares regression, generalised linear regression, ordered logit, negative binomial, multiple classification analysis (MCA), and artificial neural networks. These models typically employ explanatory variables such as employment, industry category, gross floor area, commodity type and years in business (Balla et al, 2022). However, a lack of establishment-level data on firm characteristics and actual flows has hindered the development of such models for urban freight systems, especially in developing countries (Sahu & Pani, 2019).…”