arXiv subjects

Yang Yuan

Publications and source records attributed to Yang Yuan.

59 records · Page 4Linked to original sources

arXiv2016

Kulfi: Robust Traffic Engineering Using Semi-Oblivious Routing

Wide-area network traffic engineering enables network operators to reduce congestion and improve utilization by balancing load across multiple paths. Current approaches to traffic engineering can be modeled in terms of a routing component that computes forwarding paths, and a load balancing component that maps incoming flows onto those paths dynamically, adjusting sending rates to fit current conditions. Unfortunately, existing systems rely on simple strategies for one or both of these components, which leads to poor performance or requires making frequent updates to forwarding paths, significantly increasing management complexity. This paper explores a different approach based on semi-oblivious routing, a natural extension of oblivious routing in which the system computes a diverse set of paths independent of demands, but also dynamically adapts sending rates as conditions change. Semi-oblivious routing has a number of important advantages over competing approaches including low overhead, nearly optimal performance, and built-in protection against unexpected bursts of traffic and failures. Through in-depth simulations and a deployment on SDN hardware, we show that these benefits are robust, and hold across a wide range of topologies, demands, resource budgets, and failure scenarios.

cs.NI↗

arXiv2015

Optimal Auctions vs. Anonymous Pricing

For selling a single item to agents with independent but non-identically distributed values, the revenue optimal auction is complex. With respect to it, Hartline and Roughgarden (2009) showed that the approximation factor of the second-price auction with an anonymous reserve is between two and four. We consider the more demanding problem of approximating the revenue of the ex ante relaxation of the auction problem by posting an anonymous price (while supplies last) and prove that their worst-case ratio is e. As a corollary, the upper-bound of anonymous pricing or anonymous reserves versus the optimal auction improves from four to $e$. We conclude that, up to an $e$ factor, discrimination and simultaneity are unimportant for driving revenue in single-item auctions.

cs.GT↗

arXiv2015

Escaping From Saddle Points --- Online Stochastic Gradient for Tensor Decomposition

We analyze stochastic gradient descent for optimizing non-convex functions. In many cases for non-convex functions the goal is to find a reasonable local minimum, and the main concern is that gradient updates are trapped in saddle points. In this paper we identify strict saddle property for non-convex problem that allows for efficient optimization. Using this property we show that stochastic gradient descent converges to a local minimum in a polynomial number of iterations. To the best of our knowledge this is the first work that gives global convergence guarantees for stochastic gradient descent on non-convex functions with exponentially many local minima and saddle points. Our analysis can be applied to orthogonal tensor decomposition, which is widely used in learning a rich class of latent variable models. We propose a new optimization formulation for the tensor decomposition problem that has strict saddle property. As a result we get the first online algorithm for orthogonal tensor decomposition with global convergence guarantee.

cs.LG↗

arXiv2014

Simple and Near-Optimal Mechanisms For Market Intermediation

A prevalent market structure in the Internet economy consists of buyers and sellers connected by a platform (such as Amazon or eBay) that acts as an intermediary and keeps a share of the revenue of each transaction. While the optimal mechanism that maximizes the intermediary's profit in such a setting may be quite complicated, the mechanisms observed in reality are generally much simpler, e.g., applying an affine function to the price of the transaction as the intermediary's fee. Loertscher and Niedermayer [2007] initiated the study of such fee-setting mechanisms in two-sided markets, and we continue this investigation by addressing the question of when an affine fee schedule is approximately optimal for worst-case seller distribution. On one hand our work supplies non-trivial sufficient conditions on the buyer side (i.e. linearity of marginal revenue function, or MHR property of value and value minus cost distributions) under which an affine fee schedule can obtain a constant fraction of the intermediary's optimal profit for all seller distributions. On the other hand we complement our result by showing that proper affine fee-setting mechanisms (e.g. those used in eBay and Amazon selling plans) are unable to extract a constant fraction of optimal profit in the worst-case seller distribution. As subsidiary results we also show there exists a constant gap between maximum surplus and maximum revenue under the aforementioned conditions. Most of the mechanisms that we propose are also prior-independent with respect to the seller, which signifies the practical implications of our result.

cs.GT↗

arXiv2013

On the Ratio of Revenue to Welfare in Single-Parameter Mechanism Design

What fraction of the potential social surplus in an environment can be extracted by a revenue-maximizing monopolist? We investigate this problem in Bayesian single-parameter environments with independent private values. The precise answer to the question obviously depends on the particulars of the environment: the feasibility constraint and the distributions from which the bidders' private values are sampled. Rather than solving the problem in particular special cases, our work aims to provide universal lower bounds on the revenue-to-welfare ratio that hold under the most general hypotheses that allow for non-trivial such bounds. Our results can be summarized as follows. For general feasibility constraints, the revenue-to-welfare ratio is at least a constant times the inverse-square-root of the number of agents, and this is tight up to constant factors. For downward-closed feasibility constraints, the revenue-to-welfare ratio is bounded below by a constant. Both results require the bidders' distributions to satisfy hypotheses somewhat stronger than regularity; we show that the latter result cannot avoid this requirement.

cs.GT↗