Learning and Planning for Time-Varying MDPs Using Maximum Likelihood Estimation.
Distribution of the number of citations over years.