Using Support Vector Machines to Predict Global Food Price Index


Shady I. Altelbany,  Anwar A. Abualhussein
AL-Azhar University – Gaza.


This research aims to Using Support Vector Machines (SVM) to predict global food price index, during the period from January 1990 to August 2020. The SVM model, with (Cost(C) = 10000, Epsilon(ε) = 0.1, gamma(γ) =90) has the lowest value of training error, with a small number of support vectors, is the best fit for monthly food price index predicting among all other SVM models with different parameter values.