Abstract:Consider the consensus problem of minimizing f(x) = Sigma(n)(i=1) fi(x), where x is an element of R-p and each f(i) is only known to the individual agent i in a connected network of n agents. To solve this problem and obtain the solution, all the agents collaborate with their neighbors through information exchange. This type of decentralized computation does not need a fusion center, offers better network load balance, and improves data privacy. This paper studies the decentralized gradient descent method [A. Nedic and A. Ozdaglar, IEEE Trans. Automat. Control, 54 (2009), pp. 48-61], in which each agent i updates its local variable x((i)) is an element of R-n by combining the average of its neighbors' with a local negative-gradient step -alpha del f(i)(x((i))). The method is described by the iteration x((i)) (k + 1) <- Sigma(n)(j=1) w(ij)x((j))(k)-alpha del f(i)(x((i))(k)), for each agent i, where w(ij) is nonzero only if i and j are neighbors or i = j and the matrix W = [w(ij)] is an element of R-n (x) (n) is symmetric and doubly stochastic. This paper analyzes the convergence of this iteration and derives its rate of convergence under the assumption that each f(i) is proper closed convex and lower bounded, del f(i) is Lipschitz continuous with constant L-fi > 0, and the stepsize a is fixed. Provided that alpha <= min{(1 + lambda(n)) (W))/L-h, 1/L-(f) over bar}, where L-h = max(i){L-fi} and L-(f) over bar = (1)(n) Sigma(n)(i=1) L-fi, the objective errors of all the local solutions and the networkwide mean solution reduce at rates of O(1/k) until they reach a level of O(alpha). If fi are strongly convex with modulus mu(fi), and alpha <= min{(1+lambda(n)(W))/L-h, 1/(L-f+mu(f))}, where mu(f) - 1/n Sigma(n)(i=1) mu(fi), then all the local solutions and the mean solution converge to the global minimizer x* at a linear rate until reaching an O(alpha)-neighborhood of x*. We also develop an iteration for decentralized basis pursuit and establish its linear convergence to an O(alpha)-neighborhood of the true sparse signal. This analysis reveals how the convergence of x((i))(k + 1) <- Sigma(n)(j=1) wijx(j)(k) - alpha del f(i)(x((i))(k)), for each agent i, depends on the stepsize, function convexity, and network spectrum.

Decentralized Online Riemannian Optimization with Dynamic Environments

Decentralized projected Riemannian stochastic recursive momentum method for smooth optimization on compact submanifolds

Decentralized projected Riemannian gradient method for smooth optimization on compact submanifolds

Online Optimization over Riemannian Manifolds

No-regret Online Learning over Riemannian Manifolds.

Decentralized Riemannian Conjugate Gradient Method on the Stiefel Manifold

Riemannian Projection-free Online Learning

Decentralized Riemannian natural gradient methods with Kronecker-product approximations

A Riemannian Alternating Descent Ascent Algorithmic Framework for Nonconvex-Linear Minimax Problems on Riemannian Manifolds

Dynamic Regret of Distributed Online Frank-Wolfe Convex Optimization

Decentralized Weakly Convex Optimization Over the Stiefel Manifold

On the Convergence of Decentralized Gradient Descent.

Riemannian Optimistic Algorithms

Riemannian Optimization for Non-convex Euclidean Distance Geometry with Global Recovery Guarantees

Riemannian Stochastic Proximal Gradient Methods for Nonsmooth Optimization over the Stiefel Manifold

A Decentralized Proximal Gradient Tracking Algorithm for Composite Optimization on Riemannian Manifolds

Distributed Riemannian Stochastic Gradient Tracking Algorithm on the Stiefel Manifold

Online Distributed Optimization with Nonconvex Objective Functions Via Dynamic Regrets

Optimal and Efficient Algorithms for Decentralized Online Convex Optimization

Distributed Adaptive Newton Methods with Globally Superlinear Convergence

An Optimal Stochastic Algorithm for Decentralized Nonconvex Finite-sum Optimization