A random model for multidimensional fitting method

Hiba Alawieh (1) , Frédéric Bertrand (2, 3) , Myriam Maumy-Bertrand (2, 3) , Nicolas Wicker (1) , Baydaa Al Ayoubi (4)

1 LPP - Laboratoire Paul Painlevé - UMR 8524
2 Labex_IRMIA - Institut de Recherche en Mathématiques, Interactions et Applications
3 IRMA - Institut de Recherche Mathématique Avancée
4 LU / ULB - الجامعة اللبنانية [بيروت] = Lebanese University [Beirut] = Université libanaise [Beyrouth]

Multidimensional fitting (MDF) method is a multivariate data analysis method recently developed and based on the fitting of distances. Two matrices are available: one contains the coordinates of the points and the second contains the distances between the same points. The idea of MDF is to modify the coordinates through modification vectors in order to fit the new distances calculated on the modified coordinates to the given distances. In the previous works, the modification vectors are taken as deterministic variables, so here we want to take into account the random effects that can be produce during the modification. An application in the sensometric domain is also given.

Domaines: Statistiques [math.ST]; Statistiques [stat]; Machine Learning [stat.ML]; Méthodologie [stat.ME]; Applications [stat.AP]; Calcul [stat.CO]; Théorie [stat.TH]

Hiba Alawieh, Frédéric Bertrand, Myriam Maumy-Bertrand, Nicolas Wicker, Baydaa Al Ayoubi. A random model for multidimensional fitting method. 2018. ⟨hal-03577082⟩