Institutional Support for Innovations in Econometric and Statistical Computation

点击翻译按钮获取中文摘要

SBR-9422505 Paul Ruud and Daniel McFadden The use of advanced statistical and econometric techniques by economists doing empirical research has always been constrained by the computational resources available to them. For example, until the invention of the PC about twelve years ago, most economists were limited by the limited availability and high cost of mainframe computers to using ordinary least squares. In recent years the computational power available to most economists has grown so rapidly that the econometric techniques that are computationally feasible have advanced well beyond the ones used in day to day research by most practicing economists and taught to most cmllege students. There is a growing need to bring the many newly feasible econometric methods to the attention of applied researchers. Such dissemination is the purpose of this set of conferences on innovations in statistical and econometric computation. These conferences will provide a focal point for the dissemination of research, for the closer interaction of researchers in statistics and econometrics, for the demonstration of state-of-the-art software, and for instruction of graduate students. These conferences have three objectives. The first objective is to bring together statisticians and econometricians. Statisticians can learn of the methodological and computational problems facing econometricians and econometricians can learn of new methods developed by statisticians. The second objective it to foster the development of new computational algorithms. This is done through "hands-on" demonstrations of new computational methods by those who have invented them. The third objective is to introduce graduate students and junior faculty to the statistical and econometric community. The relatively small size of the conferences creates an intimate atmosphere conducive to lively interaction.