Predicting silver price by applying a coupled multiple linear regression (MLR) and imperialist competitive algorithm (ICA)

This paper seeks to estimate and predict the global price of silver as a strategic metal using a combined multiple linear regression (MLR) and imperialist competitive algorithm (ICA). For this purpose, the global silver, copper, and aluminum prices were studied during 2009-2019. Then, the global prices of silver, copper, and aluminum were considered each as one of the input parameters, and, in return, the silver price was chosen as the target parameter. Using the Table Curve 2D & 3D software, the comprehensive statistical relationships between the input and output parameters were specified and suggested. Subsequently, the SPSS v25 software and the stepwise method were used to suggest the best nonlinear regression relationship with the 85% confidence level. Eventually, the optimal coefficients of the proposed statistical relation were determined by applying the ICA, which resulted in the improving results and also the reducing prediction error up to 1%.