Cathy H. Xia

Jia Liu, Cathy H. Xia, Ness B. Shroff et al.

Due to the rapidly growing scale and heterogeneity of wireless networks, the design of distributed cross-layer optimization algorithms have received significant interest from the networking research community. So far, the standard distributed cross-layer approach in the literature is based on first-order Lagrangian dual decomposition and the subgradient method, which suffers a slow convergence rate. In this paper, we make the first known attempt to develop a distributed Newton's method, which is second-order and enjoys a quadratic convergence rate. However, due to interference in wireless networks, the Hessian matrix of the cross-layer problem has an non-separable structure. As a result, developing a distributed second-order algorithm is far more challenging than its counterpart for wireline networks. Our main results in this paper are two-fold: i) For a special network setting where all links mutually interfere, we derive decentralized closed-form expressions to compute the Hessian inverse; ii) For general wireless networks where the interference relationships are arbitrary, we propose a distributed iterative matrix splitting scheme for the Hessian inverse. These results successfully lead to a new theoretical framework for cross-layer optimization in wireless networks. More importantly, our work contributes to an exciting second-order paradigm shift in wireless networks optimization theory.

Simon Zhang, Ryan P. DeMilt, Kun Jin et al.

Out-of-distribution (OoD) generalization occurs when representation learning encounters a distribution shift. This occurs frequently in practice when training and testing data come from different environments. Covariate shift is a type of distribution shift that occurs only in the input data, while the concept distribution stays invariant. We propose RIA - Regularization for Invariance with Adversarial training, a new method for OoD generalization under convariate shift. Motivated by an analogy to $Q$-learning, it performs an adversarial exploration for training data environments. These new environments are induced by adversarial label invariant data augmentations that prevent a collapse to an in-distribution trained learner. It works with many existing OoD generalization methods for covariate shift that can be formulated as constrained optimization problems. We develop an alternating gradient descent-ascent algorithm to solve the problem, and perform extensive experiments on OoD graph classification for various kinds of synthetic and natural distribution shifts. We demonstrate that our method can achieve high accuracy compared with OoD baselines.