Remache, CherifMaamri, RamdaneZaidi, Sahnoun2022-04-272022-04-272013978-1-4799-3343-3http://hdl.handle.net/123456789/13018The use of large number of variables to describe complex systems generates redundant elements and an excessive complexity with a poor transparency of the model describing the system. A particular interest is necessary to select relevant input variables to identify the input structure. We propose a method to select this structure in order to provide a compromise between the complexity, the transparency and the numerical accuracy of the studied model.enFuzzy ruleFuzzy inference systemInterpretative ruleComplex systemsHigh-dimensional SystemComplexity reductionArticle