After half a decade of research, much has already been written on the topic of serial correlation in regression models. However, CEIBS Joint Professor of Decision Science Yue Fang outlines a new procedure for generating forecasts for regression models.

Regression models are used to make accurate predictions on the impact one variable will have on one or more other variables. Such models are widely used in academia and, in their more basic form, within the business world.

Professor Fang’s paper, written with co-author Sergio Koreisha (Philip Knight Professor of Business at the University of Oregon Lunquist College of Business), is entitled “Using Least Squares to Generate Forecasts in Regressions with Serial Correlation.” It has been accepted for publication in the Journal of Time Series Analysis.