A Joint Kriging Model with Application to Constrained Classification - FAYOL / DEMO : Décision en Entreprise : Modélisation, Optimisation Access content directly
Preprints, Working Papers, ... Year : 2023

A Joint Kriging Model with Application to Constrained Classification


Interpolating or predicting data is of utmost importance in machine learning, and Gaussian Process Regression is one of the numerous techniques that is often used in practice. This paper considers the case of multi-input and multi-output data. It proposes a simple Joint Kriging model where common combination weights are applied to all output variables at the same time. This dramatically reduces the number of hyperparameters to be optimized, while keeping nice interpolating properties. An original constraint on predicted values is also introduced, useful for considering external information or adverse scenarios. Finally, it is shown that applied to membership degrees, the model is especially helpful for fuzzy classification problems. In particular, the model allows for prescribed average percentages of each class in predictions. Numerical illustrations are provided for both simulated and real data, and show the importance of the constraint on predicted values. The method also competes with state-of-the-art techniques on an open real world data set.
Fichier principal
Vignette du fichier
JointKriging-Hal.pdf (1.04 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04208454 , version 1 (15-09-2023)
hal-04208454 , version 2 (27-09-2023)


  • HAL Id : hal-04208454 , version 2


Didier Rullière, Marc Grossouvre. A Joint Kriging Model with Application to Constrained Classification. 2023. ⟨hal-04208454v2⟩
69 View
75 Download


Gmail Mastodon Facebook X LinkedIn More