The paper asks the question - as time series analysis moves from consideration of conditional mean values and variances to unconditional distributions, do some of the familiar concepts devised for the first two moments continue to be helpful in the more general area? Most seem to generalize...

A new method, called Relevant Transformation of the Inputs Network Approach is proposed as a tool for model building. It is designed around flexibility (with nonlinear transformations of the predictors of interest), selective search within the range of possible models, out-of-sample forecasting...