Description
Due to the limited number of events available and finite resolution of detectors, unfolding measurements of neutrino interaction cross sections is a statistically ill-posed problem. To address this, T2K analyses employ several different methods of regularisation. A particular effort has been made to mitigate the potential bias associated with choosing the termination of Expectation-Maximisation iterations (often called D'Agostini iterations) based on mock-data studies by moving to more data-driven techniques (such as the L-Curve scan). In this contribution we present the current analysis techniques alongside planned improvements with a view to receiving feedback from the particle physics statistics community
