In many applications, it is of interest to identify anomalous behavior within a dynamic interacting system. Such anomalous interactions are reflected by structural changes in the network representation of the system. We propose and investigate the use of the degree corrected stochastic block model (DCSBM) to model and monitor dynamic networks that undergo a significant structural change. We apply statistical process monitoring techniques to the estimated parameters of the DCSBM to identify significant structural changes in the network. We apply our surveillance strategy to a dynamic US Senate covoting network. We detect significant changes in the political network that reflect both times of cohesion and times of polarization among Republican and Democratic party members. Our analysis demonstrates that the DCSBM monitoring procedure effectively detects local and global structural changes in complex networks, providing useful insights into the modeled system. The DCSBM approach is an example of a general framework that combines parametric random graph models and statistical process monitoring techniques for network surveillance.

In this expository paper we give an overview of some statistical methods for the monitoring of social networks. We discuss the advantages and limitations of various methods as well as some relevant issues. One of our primary contributions is to give the relationships between network monitoring methods and monitoring methods in engineering statistics and public health surveillance. We encourage researchers in the industrial process monitoring area to work on developing and comparing the performance of social network monitoring methods. We also discuss some of the issues in social network monitoring and give a number of research ideas.

Low-energy fractures of the proximal humerus indicate osteoporosis and it is important to direct treatment to this group of patients who are at high risk of further fracture. Data were prospectively collected from 79 patients (11 men, 68 women) with a mean age of 69 years (55 to 86) with fractures of the proximal humerus in order to determine if current guidelines on the measurement of the bone mineral density at the hip and lumbar spine were adequate to stratify the risk and to guide the treatment of osteoporosis. Bone mineral density measurements were made by dual-energy x-ray absorptiometry at the proximal femur, lumbar spine (L2-4) and contralateral distal radius, and the T-scores were generated for comparison. Data were also collected on the use of steroids, smoking, the use of alcohol, hand dominance and comorbidity. The mean T-score for the distal radius was -2.97 (SD 1.56) compared with -1.61 (SD 1.62) for the lumbar spine and -1.78 (SD 1.33) for the femur. There was a significant difference between the mean lumbar and radial T scores (1.36 (1.03 to 1.68); p < 0.001) and between the mean femoral and radial T-scores (1.18 (0.92 to 1.44); p < 0.001). The inclusion of all three sites in the determination of the T-score increased the sensitivity to 66% compared with that of 46% when only the proximal femur and lumbar spine were used. This difference between measurements in the upper limb compared with the axial skeleton and lower limb suggests that basing risk assessment and treatment on only the bone mineral density taken at the hip or lumbar spine may misrepresent the extent of osteoporosis in the upper limb and the subsequent risk of fracture at this site. The assessment of osteoporosis must include measurement of the bone mineral density at the distal radius to avoid underestimation of osteoporosis in the upper limb.

PACS software has many advantages, but when using systems that can display angle measurements to one-tenth of a degree caution must be exercised to ensure that reliability of these measurements is not overestimated. We found that in the context of measuring the NSA of the proximal femur the reliability of the measurement, even under the best conditions, is only ±6° for different observers.

A common and important problem arising in the study of networks is how to divide the vertices of a given network into one or more groups, called communities, in such a way that vertices of the same community are more interconnected than vertices belonging to different ones. We propose and investigate a testing based community detection procedure called Extraction of Statistically Significant Communities (ESSC). The ESSC procedure is based on p-values for the strength of connection between a single vertex and a set of vertices under a reference distribution derived from a conditional configuration network model. The procedure automatically selects both the number of communities in the network and their size. Moreover, ESSC can handle overlapping communities and, unlike the majority of existing methods, identifies "background" vertices that do not belong to a well-defined community. The method has only one parameter, which controls the stringency of the hypothesis tests. We investigate the performance and potential use of ESSC and compare it with a number of existing methods, through a validation study using four real network data sets. In addition, we carry out a simulation study to assess the effectiveness of ESSC in networks with various types of community structure, including networks with overlapping communities and those with background vertices. These results suggest that ESSC is an effective exploratory tool for the discovery of relevant community structure in complex network systems. Data and software are available at http://www.unc.edu/~jameswd/research. html.Received September 2013; revised May 2014. 1 Supported in part by NSF Grants DMS-09-07177, DMS-13-10002, DMS-06-45369, DMS-11-05581 and SES-1357622. 1. Introduction. The study of networks has been motivated by, and made significant contributions to, the modeling and understanding of complex systems. Networks are used to model the relational structure between individual units of an observed system. In the network setting, vertices represent the units of the system and edges are placed between vertices that are related in some way. Network-based models have been used in a variety of disciplines: in biology to model protein-protein and gene-gene interactions; in sociology to model friendship and information flow among a group of individuals; and in neuroscience to model the relationship between the organization and function of the brain. In many of these applications, the vertices of the network under study can naturally be subdivided into communities. Informally, a community is a group of vertices that are more connected to each other than they are to the remainder of the network. More rigorous definitions quantify this notion of differential connection in different ways. Figure 1 illustrates a network with three disjoint communities.The problem of dividing the vertices of a given network into well-defined communities is known as community detection. Community detection has become increasingly popular, as communities have been found to identify import...

Increasing antibiotic resistance among uropathogenic Escherichia coli (UPEC) is driving interest in therapeutic targeting of nonconserved virulence factor (VF) genes. The ability to formulate efficacious combinations of antivirulence agents requires an improved understanding of how UPEC deploy these genes. To identify clinically relevant VF combinations, we applied contemporary network analysis and biclustering algorithms to VF profiles from a large, previously characterized inpatient clinical cohort. These mathematical approaches identified four stereotypical VF combinations with distinctive relationships to antibiotic resistance and patient sex that are independent of traditional phylogenetic grouping. Targeting resistance- or sex-associated VFs based upon these contemporary mathematical approaches may facilitate individualized anti-infective therapies and identify synergistic VF combinations in bacterial pathogens.

The term network surveillance is defined in general terms and illustrated with many examples. Statistical methodologies that can be used as tools for network surveillance are discussed. Details for 3 illustrative examples that address network security, surveillance for data network failures, and surveillance of email traffic flows are presented. Some open areas of research are identified.

No abstract