The near detector complex at the DUNE long-baseline neutrino oscillation experiment will use the PRISM approach: Some of the near detectors will be movable to different off-beam-axis angles to measure neutrino interaction event rates in different neutrino flux distributions. A linear combination of the event rate distributions at the different angles can then be used to predict the event rates at the far detector for different assumed values of oscillation parameters, with a reduced dependence on (and thus reduced systematic uncertainty caused by) neutrino-interaction models. The event rates measured at different angles can also be used to measure neutrino cross section as a function of neutrino energy, without having to rely on interaction models to reconstruct the energy on an event-by-event basis. This is a novel statistical approach that promises to allow measurements that simply were not possible to do in a model-independent way before. While it should work in principle, many question regarding the practical application of this methods remain and need to be investigated. A thesis in this topic would involve studying this method using simulated data, especially in regard to the question how it performs under realistic conditions.