Probabilistic Interval Predictor Based on Dissimilarity Functions

This work presents a new methodology to obtain probabilistic interval predictions of dynamical system. The proposed strategy uses stored past system measurements estimate the future evolution method relies on use dissimilarity functions conditional probability density function outputs. A family empirical functions, parameterized by means two scalars, is introduced. It shown that encompasses multivariable normal as particular case. We show presented approach constitutes generalization classical estimation methods. validation scheme used tune parameters which relies. In order prove effectiveness methodology, some numerical examples and comparisons are provided.