This paper proposes a fast and privacy preserving distributed algorithm for handling low-rank regression problems with nuclear norm constraint. Traditional projected gradient algorithms have high computation costs due to their projection steps when they are used to solve these problems. Our gossip-based algorithm, called the fast DeFW algorithm, overcomes this issue since it is projection-free. In particular, the algorithm incorporates a carefully designed decentralized power method step to reduce the complexity by distributed computation over network. Meanwhile, privacy is preserved as the agents do not exchange the private data, but only a random projection of them. We show that the fast DeFW algorithm converges for both convex and non-convex losses. As an application example, we consider the low-rank matrix completion problem and provide numerical results to support our findings.

We propose distributed algorithms for high-dimensional sparse optimization. In many applications, the parameter is sparse but high-dimensional. This is pathological for existing distributed algorithms as the latter require an information exchange stage involving transmission of the full parameter, which may not be sparse during the intermediate steps of optimization. The novelty of this work is to develop communication efficient algorithms using the stochastic Frank-Wolfe (sFW) algorithm, where the gradient computation is inexact but controllable. For star network topology, we propose an algorithm with low communication cost and establishes its convergence. The proposed algorithm is then extended to perform decentralized optimization on general network topology. Numerical experiments are conducted to verify our findings.

The matrix completion problem consists in reconstructing a matrix from a sample of entries, possibly observed with noise. A popular class of estimator, known as nuclear norm penalized estimators, are based on minimizing the sum of a data fitting term and a nuclear norm penalization. Here, we investigate the case where the noise distribution belongs to the exponential family and is sub-exponential. Our framework allows for a general sampling scheme. We first consider an estimator defined as the minimizer of the sum of a log-likelihood term and a nuclear norm penalization and prove an upper bound on the Frobenius prediction risk. The rate obtained improves on previous works on matrix completion for exponential family. When the sampling distribution is known, we propose another estimator and prove an oracle inequality \em w.r.t. the Kullback-Leibler prediction risk, which translates immediately into an upper bound on the Frobenius prediction risk. Finally, we show that all the rates obtained are minimax optimal up to a logarithmic factor.