2017 (modified: 11 Nov 2022)ICML 2017Readers: Everyone
Abstract:Motivated by the need for distributed learning and optimization algorithms with low communication cost, we study communication efficient algorithms for distributed mean estimation. Unlike previous ...