Perceptrons with polynomial post-processing

Satellite networks have great potential in providing global ubiquitous broadband communication. In this paper, we explore the capacity of both single-layered and two-layered satellite networks. Closed-form approximate expressions of network capacity are derived, which provide insights into the impact of different network parameters, such as number of orbits, number of satellites in each orbit, link bandwidth, existence of the seam, the relative position between the two layers, etc. We investigate the advantages of two-layered structure by comparing the capacity of single-layered and two-layered networks. The results show that the network capacity of two-layered networks is always no less than the total capacity of the two layers. Moreover, we obtain the network parameter condition under which the network capacity of two-layered networks is strictly larger than the sum capacity of the two layers. The results obtained in this paper can serve as a guideline for the design of efficient multi-layered satellite networks.

Pattern synchronization in a two-layer neuronal network is studied. For a single-layer network of Rulkov map neurons, there are three kinds of patterns induced by noise. Additive noise can induce ordered patterns at some intermediate noise intensities in a resonant way; however, for small and large noise intensities there exist excitable patterns and disordered patterns, respectively. For a neuronal network coupled by two single-layer networks with noise intensity differences between layers, we find that the two-layer network can achieve synchrony as the interlayer coupling strength increases. The synchronous states strongly depend on the interlayer coupling strength and the noise intensity difference between layers.

Noise models are crucial for designing image restoration algorithms, generating synthetic training data, and predicting algorithm performance. There are two related but distinct estimation scenarios. The first is model calibration, where it is assumed that the input ideal bitmap and the output of the degradation process are both known. The second is the general estimation problem, where only the image from the output of the degradation process is given. While researchers have addressed the problem of calibration of models, issues with the general estimation problems have not been addressed in the literature. In this paper, we describe a parameter estimation algorithm for a morphological, binary, page-level image degradation model. The inputs to the estimation algorithm are 1) the degraded image and 2) information regarding the font type (italic, bold, serif, sans serif). We simulate degraded images using our model and search for the optimal parameter by looking for a parameter value for which the local neighborhood pattern distributions in the simulated image and the given degraded image are most similar. The parameter space is searched using a direct search optimization algorithm. We use the p-value of the Kolmogorov-Smirnov test as the measure of similarity between the two neighborhood pattern distributions. We show results of our algorithm on degraded document images.

The consensus control problem of nonlinear cyber-physical systems with two-layer switching networks is investigated in this paper. Due to the hierarchical structure of two-layer networks, two-layer switching networks can simulate real networks more effectively than single-layer networks. In addition, the nonlinear component of the networks satisfies incremental quadratic constraints. An appropriate Lyapunov function is constructed to provide sufficient conditions for the node-to-node consensus problem. Furthermore, under some given assumptions and the node-to-node consensus, the complete consensus of two-layer networks is achieved. Finally, the validity of the method is verified by a simulation.

A node-based SIRS epidemic model on two-layer interconnected networks: Dynamical analysis of interplay between layers

In this paper we discuss the possible generalizations of a lifting theorem of a 2 × 2 2 \times 2 matrix to uniform algebras. These have applications to Hankel operators, weighted norm inequalities for conjugation operators and Toeplitz operators on uniform algebras. For example, the Helson-Szegö theorems for general uniform algebras follow.

Some comments on “The estimation of lost multi-copy documents: A new type of informetrics theory” by Egghe and Proot

In a previous paper on the allocation problem under uncertainty, Brief and Owen' propose a method for allocating costs over time that is based on the mean and variance of the distribution of future cash flows. Their allocation scheme differs from traditional methods which implicitly depend only on the best estimates, e.g., the mean values of future cash flows. Following their method the relative assignment of costs is shifted to less risky periods. Since one would expect that the uncertainty about future cash flows would increase as a function of time, we have in effect a formal mechanism for rationalizing the idea of accelerated depreciation. The work of Ijiri and Kaplan2 on a related problem also provides support for the view that conservative practice in the face of uncertainty is to accelerate the allocation of costs. In a more recent paper, Brief and Owen' restructure the allocation problem in financial accounting as part of a more general estimation problem. Their argument is as follows: Since all costs and revenues are jointly incurred over a fixed time horizon, only numbers associated with that time horizon can have economic content. Thus, periods are samples from a longer time span and measures of period performance such as the return on sales must be calculated so that they will be close to the return on

Optimized design of survivable MPLS over optical transport networks

Results using a neural network classifier for LWIR HSI target detection are presented. Detection performance using two different single layer networks with 10 and 20 nodes in the hidden layer are presented. Additionally, a two-layer network with 20 nodes in the first layer and 10 in the second is also tested. The larger networks generally performbetter, as expected, but this is not necessarily true for all targets. Neural networks significantly outperformACE across several detection metric for eight different targets. Even when ACE is applied to atmospherically-corrected emissivity datacubes, neural networks applied without atmospheric compensation still performbetter. Finally, it is found that incorporating atmospheric metadata into the network inputs can improve detection performance but is again target-dependent.

In this paper, we propose and study a two-layer network composed of a Petri net in the first layer and a ring of coupled Hindmarsh-Rose neurons in the second layer. Petri nets are appropriate platforms not only for describing sequential processes but also for modeling information circulation in complex systems. Networks of neurons, on the other hand, are commonly used to study synchronization and other forms of collective behavior. Thus, merging both frameworks into a single model promises fascinating new insights into neuronal collective behavior that is subject to changes in network connectivity. In our case, the Petri net in the first layer manages the existence of excitatory and inhibitory links among the neurons in the second layer, thereby making the chemical connections time-varying. We focus on the emergence of different types of collective behavior in the model, such as synchronization, chimeras, and solitary states, by considering different inhibitory and excitatory tokens in the Petri net. We find that the existence of only inhibitory or excitatory tokens disturbs the synchronization of electrically coupled neurons and leads toward chimera and solitary states.

We consider the robustness of multilayer networks composed of active and inactive oscillators from the viewpoint of interlayer coupling effects through the aging transition [H. Daido and K. Nakanishi, Phys. Rev. Lett. 93, 104101 (2004)]. We show in detail that two-layer networks increase or decrease their robustness depending on interlayer coupling schemes compared with single-layer networks. In addition, we find that an increase of mismatches of oscillator types (active or inactive) among interlayer-connected oscillators reduces the robustness of the networks with mean-field, chain, and diffusive interlayer couplings in two-layer networks. Moreover, we discuss the robustness of networks with more than two layers.

Considering the tedious and complex problem of solving joint probability with the bucket elimination algorithm, a parallel joint probability inference algorithm was proposed. The logical relation was analysed among each node variable and the inverse order method of the bucket elimination was used to introduce the mathematical expressions of joint probability and calculate the probabilities of one or more node failures when the network fails. The method is applied into the example of reliable analysis of electronic products and is demonstrated the simplicity and effectiveness of the method in the process of two-way Bayesian network inference. The results can be used to identify the weak links of electronic products and also provide a theoretical basis for reliability and maintenance strategies for electronic products.

This paper presents the development of several efficient learning algorithms for neural networks (ELEANNE). The ELEANNE 1 and ELEANNE 2 are two recursive least-squares learning algorithms, proposed for training single-layered neural networks with analog output. This paper also proposes a new optimization strategy for training single-layered neural networks, which provides the basis for the development of a variety of efficient learning algorithms. This optimization strategy is the source of the ELEANNE 3, a second-order learning algorithm for training single-layered neural networks with binary output. A simplified version of this algorithm, called ELEANNE 4, is also derived on the basis of some simplifying but reasonable assumptions. The two algorithms developed for single-layered neural networks provide the basis for the derivation of ELEANNE 5 and ELEANNE 6, which are proposed for training multilayered neural networks with binary output. The ELEANNE 7 is an efficient algorithm developed for training multilayered neural networks with either binary or analog output. >

A class of tests and estimators for the parameters of the Cox proportional hazards model, the accelerated failure time model, and a model for the effect of treatment on the mean of a response variable of interest are proposed that use surrogate marker data to recover information lost due to independent censoring and to adjust for bias due to dependent censoring in randomized clinical trials. We construct an adaptive test that (i) is asymptotically distribution free under the null hypothesis of no treatment effect on survival, (ii) incorporates surrogate marker data, and (iii) is guaranteed to be locally more powerful than the ordinary log-rank test against proportional hazards alternatives when the baseline failure time distribution is Weibull. The proposed test is shown to outperform the log-rank test in a series of simulation experiments. We also prove the optimal estimator within our class is semiparametric efficient by first showing that our estimation problem is a special case of the general problem of parameter estimation in an arbitrary semiparametric model with data missing at random, and then deriving a representation for the efficient score in this more general problem.