A geometric analysis of the SIRS epidemiological model on a homogeneous network

We study a fast–slow version of an SIRS epidemiological model on homogeneous graphs, obtained through the application of the moment closure method. We use GSPT to study the model, taking into account that the infection period is much shorter than the average duration of immunity. We show that the dynamics occurs through a sequence of fast and slow flows, that can be described through 2-dimensional maps that, under some assumptions, can be approximated as 1-dimensional maps. Using this method, together with numerical bifurcation tools, we show that the model can give rise to periodic solutions, differently from the corresponding model based on homogeneous mixing.

Proximal CA3 is the primary locus of age-related pattern separation deficits in rats

Age-related deficits in pattern separation have been postulated to bias the output of hippocampal memory processing toward pattern completion, which can cause deficits in accurate memory retrieval. While the CA3 region of the hippocampus is often conceptualized as a homogeneous network involved in pattern completion, growing evidence demonstrates a functional gradient in CA3 along the transverse axis, with proximal CA3 supporting pattern separation and distal CA3 supporting pattern completion. We examined the neural representations along the CA3 transverse axis in young (Y), aged memory-unimpaired (AU), and aged memory-impaired (AI) rats when different changes were made to the environment. When the environmental similarity was high (e.g., altered cues or altered environment shapes in the same room), Y and AU rats showed more orthogonalized representations in proximal CA3 than in distal CA3, consistent with prior studies showing a functional dissociation along the transverse axis of CA3. In contrast, AI rats showed less orthogonalization in proximal CA3 than Y and AU rats but showed more normal (i.e., generalized) representations in distal CA3, with little evidence of a functional gradient. When the environmental similarity was low (e.g., recordings were done in different rooms), representations in proximal and distal CA3 remapped in all rats, showing that AI rats are able to dissociate representations when inputs show greater dissimilarity. These results provide evidence that the aged-related bias towards pattern completion is due to the loss in AI rats of the normal transition from pattern separation to pattern completion along the CA3 transverse axis and, furthermore, that proximal CA3 is the primary locus of this age-related dysfunction in neural coding.

Spoofing Attack Detection Using Machine Learning in Cross-Technology Communication

Cross-technology communication (CTC) technique can realize direct communication among heterogeneous wireless devices (e.g., WiFi, ZigBee, and Bluetooth in the 2.4 G ISM band) without gateway equipment for forwarding, which makes heterogeneous wireless communication more convenient and greatly reduces communication costs. However, compared with the traditional homogeneous network model, CTC technique also makes it easier to implement spoofing attacks in heterogeneous networks. WiFi devices with long communication distances and sufficient energy supply can directly launch spoofing attacks against ZigBee devices, which brings severe security concerns for heterogeneous wireless communications. In this paper, we focus on the CTC spoofing attack, especially spoofing attacks from WiFi to ZigBee and propose a machine learning-based method to detect spoofing attacks for heterogeneous wireless networks by using physical-layer information. First, we model the received signal strength (RSS) data of legitimate ZigBee devices to construct a one-class support vector machine (OSVM) classifier for detecting CTC spoofing attacks depending on the obtained training samples. Then, we simulated CTC spoofing attacks in a live testbed and evaluated the performance of our detection method. Results show that our approach is highly effective in spoofing detection. Even if the distance between the legitimate ZigBee device and WiFi attacker is near each other (i.e., less than 2 m) and does not require a large number of samples, the detection rate and precision of our method are both over 90%. Finally, we employ the OSVM classifier to obtain samples of spoofing attacks and then explore using SVM to further improve the performance of the classifier.

Heterogeneous Graph Information Bottleneck

Most attempts on extending Graph Neural Networks (GNNs) to Heterogeneous Information Networks (HINs) implicitly take the direct assumption that the multiple homogeneous attributed networks induced by different meta-paths are complementary. The doubts about the hypothesis of complementary motivate an alternative assumption of consensus. That is, the aggregated node attributes shared by multiple homogeneous attributed networks are essential for node representations, while the specific ones in each homogeneous attributed network should be discarded. In this paper, a novel Heterogeneous Graph Information Bottleneck (HGIB) is proposed to implement the consensus hypothesis in an unsupervised manner. To this end, information bottleneck (IB) is extended to unsupervised representation learning by leveraging self-supervision strategy. Specifically, HGIB simultaneously maximizes the mutual information between one homogeneous network and the representation learned from another homogeneous network, while minimizes the mutual information between the specific information contained in one homogeneous network and the representation learned from this homogeneous network. Model analysis reveals that the two extreme cases of HGIB correspond to the supervised heterogeneous GNN and the infomax on homogeneous graph, respectively. Extensive experiments on real datasets demonstrate that the consensus-based unsupervised HGIB significantly outperforms most semi-supervised SOTA methods based on complementary assumption.

Stability analysis of the coexistence equilibrium of a balanced metapopulation model

We analyze the stability of a unique coexistence equilibrium point of a system of ordinary differential equations (ODE system) modelling the dynamics of a metapopulation, more specifically, a set of local populations inhabiting discrete habitat patches that are connected to one another through dispersal or migration. We assume that the inter-patch migrations are detailed balanced and that the patches are identical with intra-patch dynamics governed by a mean-field ODE system with a coexistence equilibrium. By making use of an appropriate Lyapunov function coupled with LaSalle’s invariance principle, we are able to show that the coexistence equilibrium point within each patch is locally asymptotically stable if the inter-patch dispersal network is heterogeneous, whereas it is neutrally stable in the case of a homogeneous network. These results provide a mathematical proof confirming the existing numerical simulations and broaden the range of networks for which they are valid.

Experimental Platform to Study Spiking Pattern Propagation in Modular Networks In Vitro

The structured organization of connectivity in neural networks is associated with highly efficient information propagation and processing in the brain, in contrast with disordered homogeneous network architectures. Using microfluidic methods, we engineered modular networks of cultures using dissociated cells with unidirectional synaptic connections formed by asymmetric microchannels. The complexity of the microchannel geometry defined the strength of the synaptic connectivity and the properties of spiking activity propagation. In this study, we developed an experimental platform to study the effects of synaptic plasticity on a network level with predefined locations of unidirectionally connected cellular assemblies using multisite extracellular electrophysiology.

Who Believes in Conspiracy Theories? Network Diversity, Political Discussion, and Conservative Conspiracy Theories on Social Media

A survey of 3,441 U.S. social media users showed that a high portion believes in conspiracy theories, and their beliefs vary widely along the party lines and socio-demographic factors. In particular, conservative conspiracy theories were more pronounced than liberal ones, and older White males with high conservatism and Protestantism showed higher endorsement of conservative conspiracy theories. Furthermore, ideological conservatives who frequently discuss politics showed higher association with a conservative conspiracy theory than conservatives who discuss politics less frequently. However, network diversity moderated the interaction of conservative ideology and political discussion such that conservatives who discuss politics frequently in a relatively heterogeneous social media network setting had lower beliefs in a conspiracy theory than conservatives who do so in a more homogeneous network.

Spreading Dynamics of a 2SIH2R, Rumor Spreading Model in the Homogeneous Network

With the rapid development of social network in recent years, the threshold of information dissemination has become lower. Most of the time, rumors, as a special kind of information, are harmful to society. And once the rumor appears, the truth will follow. Considering that the rumor and truth compete with each other like light and darkness in reality, in this paper, we study a rumor spreading model in the homogeneous network called 2SIH2R, in which there are both spreader1 (people who spread the rumor) and spreader2 (people who spread the truth). In this model, we introduced discernible mechanism and confrontation mechanism to quantify the level of people's cognitive abilities and the competition between the rumor and truth. By mean-field equations, steady-state analysis, and numerical simulations in a generated network which is closed and homogeneous, some significant results can be given: the higher the discernible rate of the rumor, the smaller the influence of the rumor; the stronger the confrontation degree of the rumor, the smaller the influence of the rumor; the larger the average degree of the network, the greater the influence of the rumor but the shorter the duration. The model and simulation results provide a quantitative reference for revealing and controlling the spread of the rumor.