Model-Aware Deep Architectures for One-Bit Compressive Variational Autoencoding. (arXiv:1911.12410v1 [eess.SP])

Parameterized mathematical models play a central role in understanding and design of complex information systems. However, they often cannot take into account the intricate interactions innate to such systems. On the contrary, purely data-driven approaches do not need explicit mathematical models for data generation and have a wider applicability at the cost of interpretability. In…

A Framework for Weighted-Sum Energy Efficiency Maximization in Wireless Networks. (arXiv:1911.12419v1 [cs.NI])

Weighted-sum energy efficiency (WSEE) is a key performance metric in heterogeneous networks, where the nodes may have different energy efficiency (EE) requirements. Nevertheless, WSEE maximization is a challenging problem due to its nonconvex sum-of-ratios form. Unlike previous work, this paper presents a systematic approach to WSEE maximization under not only power constraints, but also data…

On Optimizing Energetic Cost of Noise Reduction in Systems with Negative Feedback. (arXiv:1911.12363v1 [cond-mat.stat-mech])

Many biological functions require the dynamics to be necessarily driven out-of-equilibrium. In contrast, in various contexts, a nonequilibrium dynamics at fast timescales can be described by an effective equilibrium dynamics at a slower timescale. In this work we study the two different aspects (i) the energy-efficiency trade-off for a specific nonequilibrium linear dynamics of two…

Stablizer-Free Weak Galerkin Methods for Monotone Quasilinear Elliptic PDEs. (arXiv:1911.12390v1 [math.NA])

In this paper, we study the stablizer-free weak Galerkin methods on polytopal meshes for a class of second order elliptic boundary value problems of divergence form and with gradient nonlinearity in the principal coefficient. With certain assumptions on the nonlinear coefficient, we show that the discrete problem has a unique solution. This is achieved by…

Goodness-of-fit test for the bivariate Hermite distribution. (arXiv:1911.12400v1 [math.ST])

This paper studies the goodness of fit test for the bivariate Hermite distribution. Specifically, we propose and study a Cram\’er-von Mises-type test based on the empirical probability generation function. The bootstrap can be used to consistently estimate the null distribution of the test statistics. A simulation study investigates the goodness of the bootstrap approach for…

Dynamical fitness models: evidence of universality classes for preferential attachment graphs. (arXiv:1911.12402v1 [math.PR])

In this paper we define a family of preferential attachment models for random graphs with fitness in the following way: independently for each node, at each time step a random fitness is drawn according to the position of a moving average process with positive increments. We will define two regimes in which our graph reproduces…

Roman and Vatican Crossover Designs. (arXiv:1911.12403v1 [math.CO])

Latin squares with a balance property among adjacent pairs of symbols—being “Roman” or “row-complete”—have long been used as uniform crossover designs with the number of treatments, periods and subjects all equal. This has been generalized in two ways: to crossover designs with more subjects and to balance properties at greater distances. We consider both of…

A concept of weak Riesz energy with application to condensers with touching plates. (arXiv:1911.12406v1 [math.CA])

We proceed further with the study of minimum weak Riesz energy problems for condensers with touching plates, initiated jointly with Bent Fuglede (Potential Anal. 51 (2019), 197–217). Having now added to the analysis constraint and external source of energy, we obtain a Gauss type problem, but with weak energy involved. We establish sufficient and/or necessary…

Calibrationless Parallel MRI using Model based Deep Learning (C-MODL). (arXiv:1911.12443v1 [cs.LG])

We introduce a fast model based deep learning approach for calibrationless parallel MRI reconstruction. The proposed scheme is a non-linear generalization of structured low rank (SLR) methods that self learn linear annihilation filters from the same subject. It pre-learns non-linear annihilation relations in the Fourier domain from exemplar data. The pre-learning strategy significantly reduces the…