LearningWeights for Linear Combination of Forecasting Methods

The linear combination of forecasts is a procedure that has improved the forecasting accuracy for different time series. We present here the use of machine learning techniques to define numerical weights for the linear combination of forecasts. In this approach, a machine learning technique uses features of the series at hand to define the adequate weights for a pre-defined number of forecasting methods. In order to evaluate this solution, we implemented a prototype that uses a MLP network to combine two widespread methods. The performed experiments revealed significantly accurate forecasts.