Communication and Memory Efficient Testing of Discrete Distributions. Diakonikolas, I., Gouleakis, T., Kane, D. M., & Rao, S. In 32nd Conference on Learning Theory, COLT 2019, volume 99, of Proceedings of Machine Learning Research, pages 1070–1106, 2019. PMLR.
[DGKR19] Considers global memory constraints, in the blackboard model: the communication bound is for the total communication over all users, and as such the results are incomparable to works on local constraints, though some of the techniques can carry over. Establish tight or near-tight bounds for identity testing of discrete distributions; the techniques also apply to testing under memory constraints.

bibtex   
@inproceedings{DGKR19,
  title = 	 {Communication and Memory Efficient Testing of Discrete Distributions},
  author = 	 {Diakonikolas, Ilias and Gouleakis, Themis and Kane, Daniel M. and Rao, Sankeerth},
  booktitle = 	 {32nd Conference on Learning Theory, {COLT} 2019},
  pages = 	 {1070--1106},
  year = 	 {2019},
  volume = 	 {99},
  series = 	 {Proceedings of Machine Learning Research},
  publisher = 	 {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v99/diakonikolas19a/diakonikolas19a.pdf},
  bibbase_note = {<div class="well well-small bibbase"><span class="bluecite">[DGKR19]</span> Considers global memory constraints, in the blackboard model: the communication bound is for the total communication over all users, and as such the results are incomparable to works on local constraints, though some of the techniques can carry over. Establish tight or near-tight bounds for identity testing of discrete distributions; the techniques also apply to testing under memory constraints.</div>}
}

Downloads: 0