“…In some cases, for instance in statistics of particles (see Stoyan & Stoyan 1994) and in partially identified models in econometrics (see Molchanov & Molinari 2018), the data consist of convex sets, which is the setting of this paper. In the simplest one-dimensional case, observations are given by intervals, for example, data on daily price ranges in finance, imprecise measurements, salary brackets in econometrics, to name a few sources, see Beresteanu & Molinari (2008), Blanco-Fernández, Colubi & González-Rodríguez (2012), Blanco-Fernández, Corral & González-Rodríguez (2011), Manski & Tamer (2002) and Yang et al (2016) and references therein. A substantial body of these works focuses on regression with interval responses and sometimes also interval regressors.…”