I presented my dynamic pricing paper at NeurIPS MLxOR workshop. Under the covariate shift and local differential privacy, we use pessimistic value-based off-policy learning algorithm for continuous treatment (pricing) assignment policy. I also thank INFORMS Applied Probability Society for the travel grant!