System forecast using a case-based reasoning methodology
Guilherme Conceição Rocha, Karl Heinz Kienitz · 2011
This paper describes a case-based reasoning methodology for system forecast. For this purpose, a system condition function is defined, a time series is built and split in different scenarios, and a progression model for each scenario is determined. The forecast model for the last scenario is determined by merging the closest past scenario forecast model and the current scenario progression model.