Empirical Model Discovery and Theory Evaluation ― Automatic Selection Methods in Econometrics

Economic models of empirical phenomena are developed for a variety of reasons, themost obvious of which is the numerical characterization of available evidence, in a suitablyparsimonious form. Another is to test a theory, or evaluate it against the evidence; still anotheris to forecast future outcomes. Building such models involves a multitude of decisions, and thelarge number of features that need to be taken into account can overwhelm the researcher. Automaticmodel selection, which draws on recent advances in computation and search algorithms, can create,and then empirically investigate, a vastly wider range of possibilities than even the greatestexpert. In this book, leading econometricians David Hendry and Jurgen Doornik report on theirseveral decades of innovative research on automatic model selection. After introducing theprinciples of empirical model discovery and the role of model selection, Hendry and Doornik outlinethe stages of developing a viable model of a complicated evolving process. They discuss thediscovery stages in detail, considering both the theory of model selection and the performance ofseveral algorithms. They describe extensions to tackling outliers and multiple breaks, leading tothe general case of more candidate variables than observations. Finally, they briefly considerselecting models specifically for forecasting.

David F. Hendry is Professor of Economics and Director of the Program in Economic Modeling, Institute for New Economic Thinking at the Oxford Martin School, University of Oxford. Jurgen A. Doornik is James Martin Fellow, Institute for New Economic Thinking, at the Oxford Martin School, University of Oxford, and a Director of OxMetrics Technologies Ltd.