Learning non-monotonic preferences, a new algorithm

Capturing preferential system of the Decision Maker (DM), given a ranking of alternatives, is a challenging research question in preference disaggregation field. UTA methods are well-known in the literature, addressing this question by a linear programming model. In most of the UTA-based methods, a monotonic value function has been applied, which limits the applicability of the method. Non-monotonic UTA-based methodologies, on the other hand, are computationally intensive. In this paper we introduce a faster and simpler model, capable of learning additive non-monotonic utility functions.