Accelerated Stochastic Algorithms for Convex-Concave Saddle-Point Problems

We develop stochastic first-order primal-dual algorithms to solve a class of convex-concave saddle-point problems. When the saddle function is strongly convex in the primal variable, we develop the first stochastic restart scheme for this problem. When the gradient noises obey sub-Gaussian distributions, the oracle complexity of our restart scheme is strictly better than any of the existing methods, even in the deterministic case. Furthermore, for each problem parameter of interest, whenever the lower bound exists, the oracle complexity of our restart scheme is either optimal or nearly optimal (up to a log factor). The subroutine used in this scheme is itself a new stochastic algorithm developed for the problem where the saddle function is nonstrongly convex in the primal variable. This new algorithm, which is based on the primal-dual hybrid gradient framework, achieves the state-of-the-art oracle complexity and may be of independent interest.

Analytics of machine replacement decisions: economic life vs real options

PurposeDespite the existence of multiple asset replacement theories, the economic life replacement method remains a major practical technique for making rational machine replacement decisions. The purpose of this paper is to bridge this method with comprehensive data analytic tools and make it applicable it to modern business reality with abundant data on operating and replacement costs.Design/methodology/approachThis study employs operations research, discrete and continuous optimization, applied mathematical modeling, data analytics, industrial economics and real options theory.FindingsConstructed stochastic algorithms extend the deterministic economic life method and are compared to the contemporary theory of stochastic asset replacement based on real options and dynamic programming. It is proven that both techniques deliver similar results when the cost volatility is small. A major theoretic finding is that the cost uncertainty speeds up the replacement decision.Research limitations/implicationsThis research suggests that the proposed stochastic algorithms may become an important tool for managerial decisions about replacement of many similar machines with detailed data on operating and replacement costs.Originality/valueCompared to the real options replacement theory, major advantages of the proposed algorithms are that they work equally well for any distribution of age-dependent stochastic operating cost. The algorithms are tested on a real industrial case about replacement of medical imaging devices. Numeric simulation supports obtained analytic outcomes.