NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID: 8254 Low-Complexity Nonparametric Bayesian Online Prediction with Universal Guarantees

This paper proposes a novel online algorithm for sequential probability prediction with side information. It combines nice theoretical results with sound experiments. While the paper is generally well written, I would encourage the authors to make the paper more accessible, by providing a bit more background on the different concepts introduced, either in the main paper or in the supplementary.