Title: Assessing the predictive strength of different radial basis function interpolation techniques for diesel fuel price prediction

Authors: Paul Boye; Yao Yevenyo Ziggah

Addresses: Department of Mathematical Sciences, Faculty of Engineering, University of Mines and Technology, Ghana ' Department of Geomatic Engineering, Faculty of Mineral Resources Technology, University of Mines and Technology, Ghana

Abstract: Accurate prediction of diesel fuel price (DFP) is of great importance to the world because diesel fuel is an important commodity which has widely been used. This study for the first time assessed radial basis function (RBF) interpolation for predicting DFP. The methods evaluated were the biharmonic, multi-quadric, inverse multi-quadric, thin plate spline and Gaussian. In the study, interest and inflation rates obtained from Bank of Ghana served as the predictor variables while DFP acquired from Ghana National Petroleum Authority was the response variable. The robustness of the interpolation methods were assessed using noise to signal ratio (NSR), relative error correction (REC), mean absolute percentage error (MAPE), and relative root mean square error (RRMSE). The analytical results revealed that the biharmonic had the best results with 22.304% (MAPE), 99.613% (NSR), 77.696% (REC) and 2.239 (RRMSE). The biharmonic could be useful to policymakers, investors, and managers of the crude oil industry.

Keywords: diesel fuel price; DFP; radial basis function interpolation; Taylor diagram; macroeconomic variables.

DOI: 10.1504/IJBFMI.2021.10040752

International Journal of Business Forecasting and Marketing Intelligence, 2021 Vol.7 No.1, pp.13 - 22

Received: 11 Apr 2021
Accepted: 16 May 2021

Published online: 07 Jan 2022 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article