Application of Triple Exponential Smoothing Technique in the Analysis of Time Series with a Quadratic Trend

  • S. L. Shrestha Central Department of Statistics, Tribhuvan University Kirtipur Kathmandu


Exponential Smoothing Technique is a useful forecasting tool in time series analysis. In contrast to the Regression Approach of forecasting, the technique is based upon unequally weighting the observed data with recent observations getting higher weightage than the remote ones. One such technique is the Triple Exponential Smoothing which is applied to the time series with a quadratic trend. The paper aims to highlight this method of forecasting with reference to time series of the Annual Number of Tourist Anivals in Nepal during the period 1971-1997. The predicted series of data arising from the Triple Exponential Smoothing statistically fits to the observed time series considered.