Two modified DANE algorithms, DANE-LS with line search and DANE-HB with heavy-ball momentum, are proved to converge in fewer communication rounds than plain DANE on strongly convex objectives, including non-quadratic losses.
Communication complexity of distributed convex learning and optimization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On Convergence of Distributed Approximate Newton Methods: Globalization, Sharper Bounds and Beyond
Two modified DANE algorithms, DANE-LS with line search and DANE-HB with heavy-ball momentum, are proved to converge in fewer communication rounds than plain DANE on strongly convex objectives, including non-quadratic losses.