Comparison of Some Forecasting Techniques in Time Series Analysis

Raydonal Ospina & Bartolomeu Zamprogno

 

Abstract     

In this work, we compared the forecasting performance of three approaches: Box & Jenkins, exponential smoothing and neural networks, under different forecasting horizons. Two series are considered.

 

Key words: Box & Jenkins models, Exponential smoothing, Neural networks, Forecast.

 

PDF (Portuguese)