1: \begin{abstract}

2: 	This paper considers the decentralized optimization problem, which has applications in large scale machine learning, sensor networks, and control theory. 

3: 	We propose a novel algorithm that can achieve near optimal communication complexity, matching the known lower bound up to a logarithmic factor of the condition number of the problem.

4: 	Our theoretical results give affirmative answers to the open problem on whether there exists an algorithm that can achieve a communication complexity (nearly) matching the lower bound depending on the global condition number instead of the local one. 

5: 	Moreover, the proposed algorithm achieves the optimal computation complexity matching the lower bound up to universal constants.

6: 	%of $\cO\left(\sqrt{\kappa_{g}}\log\left(\frac{1}{\epsilon}\right)\right)$.

7: 	Furthermore, to achieve a linear convergence rate, our algorithm  \emph{doesn't} require the individual functions to be (strongly) convex.  

8: 	Our method relies on a novel combination of known techniques

9: 	including Nesterov's accelerated gradient descent,

10: 	multi-consensus and gradient-tracking. The analysis is new,

11: 	and may be applied to other related problems. Empirical studies demonstrate the effectiveness of our method for machine learning applications. 

12: \end{abstract}

13: