Machine learning approach to inverse problem and unfolding procedure

A procedure for unfolding the true distribution from experimental data is presented. Machine learning method are applied for the identification an apparatus function and solving inverse problem simultaneously. A priori information about the true distribution which is known from theory or previous experiments is used for Monte-Carlo simulation of the training sample. The stability of the result of the unfolding is obtained by a sensible binning and by application of D-optimization. The unfolding procedure may be applied for detectors with a linear or nonlinear transformation of a true distribution into the experimentally measured one. The dimensionality of the solved problem can be arbitrary.