Accelerated extra-gradient descent: a novel accelerated first-order method

Date Issued
2018Publisher Version
10.4230/LIPIcs.ITCS.2018.23Author(s)
Orecchia, Lorenzo
Diakonikolas, Jelena
Metadata
Show full item recordPermanent Link
https://hdl.handle.net/2144/38507Version
Published version
Citation (published version)
Lorenzo Orecchia, Jelena Diakonikolas. 2018. "Accelerated Extra-Gradient Descent: A Novel Accelerated First-Order Method." 9th Innovations in Theoretical Computer Science Conference (ITCS 2018). https://doi.org/10.4230/LIPIcs.ITCS.2018.23Abstract
We provide a novel accelerated first-order method that achieves the asymptotically optimal convergence rate for smooth functions in the first-order oracle model. To this day, Nesterov’s Accelerated Gradient Descent (agd) and variations thereof were the only methods achieving acceleration
in this standard blackbox model. In contrast, our algorithm is significantly different from agd,
as it relies on a predictor-corrector approach similar to that used by Mirror-Prox [18] and ExtraGradient Descent [14] in the solution of convex-concave saddle point problems. For this reason,
we dub our algorithm Accelerated Extra-Gradient Descent (axgd).
Its construction is motivated by the discretization of an accelerated continuous-time dynamics [15] using the classical method of implicit Euler discretization. Our analysis explicitly shows
the effects of discretization through a conceptually novel primal-dual viewpoint. Moreover, we
show that the method is quite general: it attains optimal convergence rates for other classes
of objectives (e.g., those with generalized smoothness properties or that are non-smooth and
Lipschitz-continuous) using the appropriate choices of step lengths. Finally, we present experiments showing that our algorithm matches the performance of Nesterov’s method, while appearing more robust to noise in some cases.
Rights
Copyright © Jelena Diakonikolas and Lorenzo Orecchia 2018; licensed under Creative Commons Attribution License (CC-BY)Collections