Pairwise kernel-based preference learning for multiple criteria decision making
Leonid M. Lyubchyk, Galina Grinberg · 2017 IEEE First Ukraine Conference on Electrical and Computer Engineering (UKRCON) · 2017
The problem of preference functions model development for multiple criteria decision-making is considered based on machine-learning approach. It is assumed that the training sample for a plurality of objects, for which decisions are made, is formed from a set of measured features or the particular criteria and the matrix of pairwise comparisons. The problem of constructing a linear preference function model is reduced to weighting coefficients estimation for features or particular criteria and is solved on basis of corresponding optimization problem with constraints. The problem of constructing a non-linear model is solved on the basis of kernel-based learning approach that allows approximating the preference function of complex structure on small training samples. For the regularization of nonlinear model preference learning procedure the proposed method of optimal model concordance is used, with estimation criteria weights are used as a priori information.