Title: Heterogeneity transforms subdiffusion into superdiffusion via ensemble self-reinforcement

Abstract: We formulate a subdiffusive master equation in continuous time with random transition probabilities such that the ensemble of random walks exhibits superdiffusion. This counter intuitive result is explained by the ensemble self-reinforcement phenomenon, which emulates conditional transition rates of an underlying discrete random walk with strong memory through a heterogeneous ensemble. We find the ensemble averaged solution of this master equation and the exact equation for the second moment, which exhibits superdiffusion. Ensemble self-reinforcement allows a smooth transition between subdiffusion and superdiffusion depending on a single parameter.