Smoothing Methods

Smoothing is a statistical method we can use to create an approximation function to remove irregularities in data and attempt to capture significant patterns. Robert Goodell Brown was the father of exponential smoothing, and in 1956 he published “Exponential Smoothing for Predicting Demand” ( ). In 1957, professor Charles C. Holt was working at CMU on forecasting and published a paper called “Forecasting Seasonals and Trends by Exponentially Weighted Moving Averages” that discussed double exponential smoothing. Three years, in 1960, a student of Holt’s, Peter R. Winters, developed an algorithm by uniting seasonality and published “Forecasting Sales by Exponentially Weighted Moving Averages” ( ).