Search

For the newly proposed definitive screening designs, we show that, among all possible run orders, the extreme values of level changes depend only on the number of factors, and the corresponding designs can be characterized through some necessary and sufficient conditions. Systematic designs with minimum or maximum level changes are constructed for practical use. [ABSTRACT FROM AUTHOR]

Definitive screening designs (DSDs) are widely used for studying quantitative factors. However, DSDs constructed from different conference matrices are not equally good. We show DSDs using Paley's conference matrices guarantee desirable performance (either optimal or near-optimal) under several important criteria. [ABSTRACT FROM AUTHOR]

Maximin distance Latin hypercube designs (LHDs) are frequently used in computer experiments, but their constructions are challenging. In this paper, we present some new results connecting maximin L 2 -distance optimality and near orthogonality for mirror-symmetric LHDs. We further propose a simple and effective method for constructing nearly orthogonal LHDs that can yield almost the largest minimum distance. The obtained designs with small and medium sizes are tabulated and their superior performances are illustrated via comparisons. [ABSTRACT FROM AUTHOR]

With the increasing popularity of personalized medicine, it is more and more crucial to capture not only the dose-effect but also the effects of the prognostic factors due to individual differences in a dose-response experiment. This paper considers the design issue for predicting semi-parametric dose-response curves in the presence of linear effects of covariates. Inspired by the Neyman-Pearson paradigm, a novel design criterion, namely bias constraint optimality, is introduced to minimize the overall prediction error. The corresponding equivalence theorems are established, the characteristics of the optimal designs are shown, and an equivalent bias compound optimality criterion is proposed for practical implementation. Based on the obtained theoretical results, efficient algorithms for searching for optimal designs are developed. Numerical simulations are given to illustrate the superior performance of the obtained optimal designs. [ABSTRACT FROM AUTHOR]

The placenta is an important organ that connects the fetus with the mother during pregnancy. Many pregnancies fail because of the placental abnormality. Nowadays, placental segmentation techniques based on magnetic resonance (MR) images are increasingly used to assist in the diagnosis of placental abnormalities. However, manual segmentation of the placenta is unreliable and time-consuming. In comparison to manually segmentation methods, the CNN-based methods perform better. However, due to the limitations of the fixed geometric structures in the traditional convolution, the features of the edge are not completely obtained. That is the challenge in placenta segmentation. By incorporating offsets into the regular grid sampling locations of standard convolution, deformable convolution allows for the utilization of spatial information to its full potential. Therefore, in this paper, we proposed a fully automatic placental segmentation method based on deformable convolution, termed as Deformable Convolution O-shape Network (DCO-Net). Specifically, we adopt the U-shaped architecture and employed a bi-directional O-shaped route by adding backward skip connections. The bi-directional O-shaped route recurrently reused the building blocks without introducing any extra parameters. To evaluate the performance of the proposed DCO-Net, a total of 871 MR images from 20 pregnant women were used for the experiments. The dice and average symmetric surface distance (ASSD) obtained by DCO-Net were 0.8488 and 3.0976, respectively. The experimental results demonstrate that the proposed DCO-Net is more efficient for automatic placental segmentation compared with the other competitive approaches. [ABSTRACT FROM AUTHOR]

Based on nonregular orthogonal arrays via the Addelman–Kempthorne method, we propose a novel construction of strong orthogonal arrays of strength 2 +. The resulting designs have 2 s n runs with s being any odd prime power and fill some run-size gaps left by existing constructions. [ABSTRACT FROM AUTHOR]

As a powerful tool for machine learning on the graph, network embedding, which projects nodes into low-dimensional spaces, has a variety of applications on complex networks. Most current methods and models are not suitable for bipartite networks, which have two different types of nodes and there are no links between nodes of the same type. Furthermore, the only existing methods for bipartite network embedding ignore the internal mechanism and highly nonlinear structures of links. Therefore, in this paper, we propose a new deep learning method to learn the node embedding for bipartite networks based on the widely used autoencoder framework. Moreover, we carefully devise a node-level triplet including two types of nodes to assign the embedding by integrating the local and global structures. Meanwhile, we apply the variational autoencoder (VAE), a deep generation model with natural advantages in data generation and reconstruction, to enhance the node embedding for the highly nonlinear relationships between nodes and complex features. Experiments on some widely used datasets show the effectiveness of the proposed model and corresponding algorithm compared with some baseline network (and bipartite) embedding techniques. [ABSTRACT FROM AUTHOR]

Abstract: Anomaly detection is indispensable for satisfying security services in mobile ad hoc network (MANET) applications. Often, however, a highly secure mechanism consumes a large amount of network resources, resulting in network performance degradation. To shift intrusion detection from existing security-centric design approaches to network performance centric design schemes, this paper presents a framework for designing an energy-aware and self-adaptive anomaly detection scheme for resource constrained MANETs. The scheme uses network tomography, a new technique for studying internal link performance based solely on end-to-end measurements. With the support of a module comprising a novel spatial-time model to identify the MANET topology, an energy-aware algorithm to sponsor system service, a method based on the expectation maximum to infer delay distribution, and a Self-organizing Map (SOM) neural network solution to profile link activity, the proposed system is capable of detecting link anomalies and localizing malicious nodes. Consequently, the proposed scheme offers a trade-off between overall network security and network performance, without causing any heavy network overload. Moreover, it provides an additional approach to monitor the spatial-time behavior of MANETs, including network topology, link performance and network security. The effectiveness of the proposed schemes is verified through extensive experiments. [Copyright &y& Elsevier]

Abstract: When simulating a mobile ad hoc network (MANET), it is important to use a realistic mobility model to reflect the actual performance of a mobile system. The spatial distribution of node locations in a mobile model plays a key role when investigating the characteristics of a MANET. However, most existing mobility models with random and simple straight line movement lead to unrealistic scenarios and non-uniform distributions, and can not describe the actual movement of Unmanned Aerial Vehicles (UAVs) connected via a MANET. To address this issue, a novel mobility model based on semi-random circular movement (SRCM) is presented. The approximate node distribution function in SRCM is derived within a 2D disk region. The relationship between application performance and node distribution is investigated for a UAV MANET, with focus on scan coverage and network connectivity. A simulation using the NS2 tool is conducted. It is shown that the presented model with a uniform distribution performs better than the popular Random Waypoint mobility model. The SRCM model with the NS2 simulator provides a realistic way for simulation and performance evaluation of UAV MANETs. [Copyright &y& Elsevier]