Master of Science
John M. Carroll
Five forcasting models, first, second, and third order exponential smoothing, moving average, and linear regression, are applied to eighty-one time series and the resultant minumum forecast erro for each of the five methods on each other the time series is computed.
Fiddleman, Richard H., "A comparison and evaluation of five forecasting techniques when applied to various stationary time series" (1966). Theses and Dissertations. 3491.