Universal Data Forecasting with an Adaptive Approach and Seasonal Technique
Phayung Meesad, Tong Srikhacha · 2006
The main component of observation data includes both trend and seasonal effects. The represented equations of forecasting models like ARIMA seem to have too many explained parameters when we need more accuracy in time series prediction. To apply these elaborate and beautifully crafted techniques we require an advanced level of knowledge and sophistication only available from specialists. However, it is more suitable if one who does not familiar with complex forecasting models can use a simple equation like applied exponential smoothing model for forecasting. We propose a simple suitable model that can be applied to most kinds of data observation types with good prediction outcome. The proposed model can be applied to calculate in a simple spreadsheet which yields good short term prediction with low error rate.