Econometric Combined with Neural Network for Coffee Price Forecasting
Econometric Combined with Neural Network for Coffee Price Forecasting

Summary/Abstract: Econometric is commonly used in economic forecasting model for short term, long term and it is applied in many different fields. While, neural network models have the advantage for fast calculating and high accuracy due to weight adjustment after training steps. The combination of econometric and neural network models will increase the effective individual models. It can achieve advantages of both econometric and neural network. In this paper, we present a method that combined two models for coffee price forecasting. Experimental results are predicted daily with the number of 427 training samples from the last 1 years with actual data with a mean deviation of 0.00243 indicating that our predictive model has acceptable accuracy.

• Details
• Contents