Section: New Results

Learning Constraint Models

Participants : Nicolas Beldiceanu, Naina Razakarison.

• In the context of learning parametrized constraint models for highly structured problems we address in [38] the problem of finding the coefficients of polynomials in several variables from example parameter and function values.

• In the context of semi structured time series where the structural aspect is related to technological constraints we deal in [24] with the problem of extracting functional dependency constraints. The problem is motivated by extracting constraints from electricity production time series and is characterized by a larger set of samples (from 7 years and from 300 plants).