Abstract
Predicting the price of gold is always an attractive area for researchers and predictors willing to define its upcoming value in a more accurate and efficient way. In this study, a Chebyshev Polynomial Neural Network (CSPNN) based predicting model is suggested for the prediction of gold price. Evolutionary approaches like Particle Swarm Optimization (PSO) and Differential Evolution (DE) are used in training of the CSPNN to derive optimally tuned weights of the network. The efficiency of CSPNN model is evaluated by means of different error measures on UK/USD and MOS/USD gold price datasets. The conclusion analyzes the results to suggest a better prediction.
