Methods and Problems of Prediction
Joachim Krauth · Neuropsychobiology · 2008
A review is given of concepts and problems in statistical prediction analysis. First some concepts of prediction (Bayesian, non-Bayesian, time series and others) are introduced. Then in particular prediction by means of multiple linear regression and multiple discrimination is discussed and hints are given with respect to alternative procedures. Various results concerning prediction in time series and by means of contingency tables and cluster analyses are reported.