Performance of Nonlinear Autoregressive Neural Networks (Narnet) for Analysis Prices in Tobacco Production and Trade
The consumer price index, (CPI) is an essential indicator that measures the inflation and the average changes in the prices of main group of products and services paid by the householders. And tobacco prices are one of the main group having an essential impact on inflation. On the one hand, economies try to control the price of tobacco products for reducing the consumption, on the other hand, the prediction of tobacco prices is highly critical for tobacco trade, production and delivery services. It also regulates the tobacco market to stimulate trade more accurately. The aim of this study is to predict the price of tobacco and smoking products in the U.S. covering January 1996 and February 2022 using the Nonlinear Autoregressive Neural Network (NARNET) method.
Empirical results show that the NARNET model with 15 neurons in the hidden layer and 3 times delays yielded higher accuracy to forecast the tobacco prices in the USA between January 1996 and February 2022.
Keywords – Nonlinear Autoregressive Neural Network (NARNET), Tobacco Price, Tobacco Trade