Search

Graph neural networks (GNNs) excel in modeling relational data such as biological, social, and transportation networks, but the underpinnings of their success are not well understood. Traditional complexity measures from statistical learning theory fail to account for observed phenomena like the double descent or the impact of relational semantics on generalization error. Motivated by experimental observations of "transductive" double descent in key networks and datasets, we use analytical tools from statistical physics and random matrix theory to precisely characterize generalization in simple graph convolution networks on the contextual stochastic block model. Our results illuminate the nuances of learning on homophilic versus heterophilic data and predict double descent whose existence in GNNs has been questioned by recent work. We show how risk is shaped by the interplay between the graph noise, feature noise, and the number of training labels. Our findings apply beyond stylized models, capturing qualitative trends in real-world GNNs and datasets. As a case in point, we use our analytic insights to improve performance of state-of-the-art graph convolution networks on heterophilic datasets. [ABSTRACT FROM AUTHOR]

Objective: To investigate the relationship of mild-to-moderate simple bilateral fetal ventriculomegaly with postnatal neurological development. Methods: Cases of simple lateral ventricular dilatation (273) were divided into bilateral mild dilatation (10.0–12.0 mm, 62), bilateral moderate dilatation (12.1–15.0 mm, 29), unilateral mild dilatation (133), and unilateral moderate dilatation (49) groups. The control group comprised 50 normal fetuses. Neurological development was assessed using Gesell Developmental Schedules (GDS) at postnatal 3, 6, 12, and 18 months. Results: At postnatal 6, 12, and 18 months, the GDS score was higher for bilateral than for unilateral dilatation (p < .05). At postnatal 3 and 6 months, the GDS score was higher for the bilateral dilatation groups than for the control group (p < .05). At postnatal 6, 12, and 18 months, the GDS score was higher for the bilateral moderate dilatation group than for the unilateral moderate dilatation group (p < .05). Further, at postnatal 3, 6, 12, and 18 months, the GDS score was higher for the bilateral moderate dilatation group than for the control group (p < .05). Conclusion: At postnatal month 6, the GDS evaluation results of the bilateral dilatation groups were significantly inferior compared to those of the unilateral dilatation group. [ABSTRACT FROM AUTHOR]

Temporality is an essential characteristic of many real-world networks and dramatically affects the spreading dynamics on networks. In this paper, we propose an information spreading model on temporal networks with heterogeneous populations. Individuals are divided into activists and bigots to describe the willingness to accept the information. Through a developed discrete Markov chain approach and extensive numerical simulations, we discuss the phase diagram of the model and the effects of network temporality. From the phase diagram, we find that the outbreak phase transition is continuous when bigots are relatively rare, and a hysteresis loop emerges when there are a sufficient number of bigots. The network temporality does not qualitatively alter the phase diagram. However, we find that the network temporality affects the spreading outbreak size by either promoting or suppressing, which relies on the heterogeneities of population and of degree distribution. Specifically, in networks with homogeneous and weak heterogeneous degree distribution, the network temporality suppresses (promotes) the information spreading for small (large) values of information transmission probability. In networks with strong heterogeneous degree distribution, the network temporality always promotes the information spreading when activists dominate the population, or there are relatively fewer activists. Finally, we also find the optimal network evolution scale, under which the network information spreading is maximized. [ABSTRACT FROM AUTHOR]

The degree-degree correlation is important in understanding the structural organization of a network and dynamics upon a network. Such correlation is usually measured by the assortativity coefficient r, with natural bounds r ∈ [-1,1]. For scale-free networks with power-law degree distribution p(k) ~ k-c, we analytically obtain the lower bound of assortativity coefficient in the limit of large network size, which is not -1 but dependent on the power-law exponent c. This work challenges the validation of the assortativity coefficient in heterogeneous networks, suggesting that one cannot judge whether a network is positively or negatively correlated just by looking at its assortativity coefficient alone. [ABSTRACT FROM AUTHOR]

The organization of real networks usually embodies both regularities and irregularities, and, in principle, the former can be modeled. The extent to which the formation of a network can be explained coincides with our ability to predict missing links. To understand network organization, we should be able to estimate link predictability. We assume that the regularity of a network is reflected in the consistency of structural features before and after a random removal of a small set of links. Based on the perturbation of the adjacency matrix, we propose a universal structural consistency index that is free of prior knowledge of network organization. Extensive experiments on disparate realworld networks demonstrate that (i) structural consistency is a good estimation of link predictability and (ii) a derivative algorithm outperforms state-of-the-art link prediction methods in both accuracy and robustness. This analysis has further applications in evaluating link prediction algorithms and monitoring sudden changes in evolving network mechanisms. It will provide unique fundamental insights into the above-mentioned academic research fields, and will foster the development of advanced information filtering technologies of interest to information technology practitioners. [ABSTRACT FROM AUTHOR]

Real-world systems, ranging from social and biological to infrastructural, can be modeled by multilayer networks. Promoting spreading dynamics in multilayer networks may significantly facilitate electronic advertising and predicting popular scientific publications. In this study, we propose a strategy for promoting the spreading dynamics of the susceptible-infected-susceptible model by adding one interconnecting edge between two isolated networks. By applying a perturbation method to the discrete Markovian chain approach, we derive an index that estimates the spreading prevalence in the interconnected network. The index can be interpreted as a variant of Katz centrality, where the adjacency matrix is replaced by a weighted matrix with weights depending on the dynamical information of the spreading process. Edges that are less infected at one end and its neighborhood but highly infected at the other will have larger weights. We verify the effectiveness of the proposed strategy on small networks by exhaustively examining all latent edges and demonstrate that performance is optimal or near-optimal. For large synthetic and real-world networks, the proposed method always outperforms other static strategies such as connecting nodes with the highest degree or eigenvector centrality. [ABSTRACT FROM AUTHOR]