The Hybrid Modern Network Model: A Multi-Technique Framework for Comprehensive Network Analysis

Recently, network analysis has been intensively investigated in several fields of science. Link prediction is a problem of predicting the existence of a link between two entities based on observed links, and it is one of the popular link mining tasks. Although many link prediction methods have been proposed, they have their merits and demerits. In this paper, we present two topics as follows: 1) In order to obtain the strategies of selecting the best link prediction methods, we perform experiments of six link prediction methods (Common Neighbors (CN) , Jaccard's Coefficient (JC) , Adamic/Adar (AA) , Shortest Path (SP) , Preferential Attachment (PA) and Hierarchical Random Graph (HRG) ) for 39 real networks. 2) We propose a new similarity that is the summation of similarities based on the logistic regression. We used 10-fold cross validation and bagging for model selection of proposed method. We estimate the accuracy and computation time of HRG, proposed method (bagging) and proposed method (10-fold cross validation) for 28 data sets. As a result of 1) , CN, JC and AA achieve good performance for the networks that has higher clustering coefficient than 0.4. SP achieves good performance for the network that has higher average shortest path length than 3. PA underperforms the random predictor for the network has lower variance of degrees than 0.5. HRG performs consistently well. As a result of 2) , accuracy of proposed methods (both of bagging and 10-fold cross validation) are reached higher than the accuracy of HRG for 17 data sets and finishes the calculation faster than HRG. Proposed methods perform good accuracy for social network, citation network, dictionary network, biological network and transfer network (journey). Proposed methods underperform for trade network, circuit network, and food web network. Sometimes, proposed method (bagging) reaches higher accuracy than the accuracy of proposed method (10-fold cross validation). Proposed method (10-fold cross validation) finishes the calculation faster than proposed method (bagging). In conclusion, proposed methods finish the calculation faster than HRG and accuracy of proposed methods reaches higher than HRG.

Today’s network is notorious for its complexity and uncertainty. Network operators often rely on network models to achieve efficient network planning, operation, and optimization. The network model is responsible for understanding the complex relationships between the network performance metrics (e.g., latency) and the network characteristics (e.g., traffic). However, we still lack a systematic approach to developing accurate and lightweight network models that are aware of the impact of network configurations (i.e., expressiveness) and provide fine-grained flow-level temporal predictions (i.e., granularity).In this paper, we propose xNet, a data-driven network modeling framework based on graph neural networks (GNN). Unlike the previous proposals, xNet is not a dedicated network model designed for specific network scenarios with constraint considerations. On the contrary, xNet provides a general approach to modeling the network characteristics of concern with relation graph representations and configurable GNN blocks. xNet learns the state transition function between time steps and rolls it out to obtain the full fine-grained prediction trajectory. We implement and instantiate xNet with three use cases. The experiment results show that xNet can accurately predict different performance metrics while achieving over two orders of magnitude of speedup compared with the conventional packet-level simulator.

Generative AI is redefining precision medicine by enabling the rational design of patient‐specific therapeutics and overcoming the inherent inefficiencies of conventional drug discovery workflows, including protracted timelines, high attrition rates, and prohibitive R&D expenditures. Generative models such as Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), transformer‐based architectures, and denoising diffusion models (DDMs) enable de novo molecular generation guided by multi‐omics datasets (genomics, transcriptomics, proteomics), facilitating the design of small molecules, peptides, and biologics tailored to individual molecular profiles. These architectures operate in high‐dimensional latent chemical spaces, allowing chemical morphing, scaffold hopping, and latent space optimization to enhance potency, selectivity, and ADME/Tox properties. Integration with QSAR models, molecular docking, molecular dynamics simulations, and protein–ligand binding affinity predictors strengthens the accuracy of drug–target interaction profiling. Reinforcement learning (RL) and graph neural network (GNN)‐based generative models further optimize lead compounds through iterative reward‐driven refinement, while SE(3)‐equivariant neural networks enable faithful 3D molecular generation and conformational stability predictions. Despite algorithmic advances, experimental validation remains indispensable to address inter‐patient metabolic heterogeneity, polypharmacology, and off‐target liabilities. Nevertheless, the convergence of generative AI with multi‐omics and high‐throughput screening platforms accelerates personalized drug discovery pipelines, establishing a paradigm shift toward precision, scalability, and translational efficiency.

The Science of Deep Learning emerged from courses taught by the author that have provided thousands of students with training and experience for their academic studies, and prepared them for careers in deep learning, machine learning, and artificial intelligence in top companies in industry and academia. The book begins by covering the foundations of deep learning, followed by key deep learning architectures. Subsequent parts on generative models and reinforcement learning may be used as part of a deep learning course or as part of a course on each topic. The book includes state-of-the-art topics such as Transformers, graph neural networks, variational autoencoders, and deep reinforcement learning, with a broad range of applications. The appendices provide equations for computing gradients in backpropagation and optimization, and best practices in scientific writing and reviewing. The text presents an up-to-date guide to the field built upon clear visualizations using a unified notation and equations, lowering the barrier to entry for the reader. The accompanying website provides complementary code and hundreds of exercises with solutions.

Cultures of the Central Highlands, New Guinea

Infer-AVAE: An attribute inference model based on adversarial variational autoencoder

Data publishing for large-scale social network has the risk of privacy leakage. Trying to solve this problem, a differential private social network data publishing algorithm named DP-HRG is proposed in the paper, which is based on Hierarchical Random Graph (HRG). Firstly, the social network is divided into 1-neighborhood subgraphs, and the HRG of each subgraph is extracted by using both Markov Monte Carlo (MCMC) and exponential mechanism to compose the HRG candidate set. Then an average edge matrix is obtained based on the HRG candidate set and perturbed by a random matrix. Finally, according to the perturbed average edge matrix, a 1-neighborhood graph is regenerated and pasted into the original social network for publishing. Experimental results show that the proposed algorithm preserves good network characteristics and better data utility while satisfying the requirement of privacy protection.

In today’s data-driven world, the accurate and consistent identification of clients across multiple platforms is a critical challenge for financial institutions, government comptroller departments, and third-party service providers. Fragmented and inconsistent data across various systems pose significant risks, including regulatory non-compliance, fraud, and operational inefficiencies. This thesis presents the development of an innovative avatar-based framework for unified client identification in banking systems, leveraging the power of Generative AI and Graph Neural Networks (GNNs). The framework synthesizes disparate client data into a single, cohesive representation in the virtual world, effectively addressing the challenges of data fragmentation and inconsistency. Generative AI models, such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs), are employed to enhance data quality by generating realistic synthetic data and imputing missing values. These models augment the existing datasets, ensuring completeness and accuracy of client profiles. Simultaneously, GNNs are utilized to model the complex relationships and interactions within the client data, capturing intricate dependencies and enhancing the accuracy of client identification. The proposed framework offers substantial benefits, including improved regulatory compliance, enhanced operational efficiency, and superior customer experience. By providing a unified view of client data, financial institutions can better detect and prevent fraudulent activities, meet stringent regulatory requirements, and deliver personalized services. Government comptroller departments can ensure more effective public fund management and maintain transparency. Third-party service providers can leverage accurate client profiles for better service delivery and risk management. The race to implement such advanced frameworks is a pivotal factor in determining leadership in the financial and administrative sectors. Institutions that adopt and integrate this technology swiftly will gain a significant competitive advantage, setting new standards in client identification and data management. This research underscores the transformative potential of combining Generative AI and GNNs in creating a robust, scalable, and efficient system for unified client identification, paving the way for future advancements in this critical field.

Complex networks of material flow in manufacturing and logistics: Modeling, analysis, and prediction using stochastic block models

Most existing graph neural networks (GNNs) learn node embeddings using the framework of message passing and aggregation. Such GNNs are incapable of learning relative positions between graph nodes within a graph. To empower GNNs with the awareness of node positions, some nodes are set as anchors. Then, using the distances from a node to the anchors, GNNs can infer relative positions between nodes. However, position-aware GNNs (P-GNNs) arbitrarily select anchors, leading to compromising position awareness and feature extraction. To eliminate this compromise, we demonstrate that selecting evenly distributed and asymmetric anchors is essential. On the other hand, we show that choosing anchors that can aggregate embeddings of all the nodes within a graph is NP-complete. Therefore, devising efficient optimal algorithms in a deterministic approach is practically not feasible. To ensure position awareness and bypass NP-completeness, we propose position-sensing GNNs (PSGNNs), learning how to choose anchors in a backpropagatable fashion. Experiments verify the effectiveness of PSGNNs against state-of-the-art GNNs, substantially improving performance on various synthetic and real-world graph datasets while enjoying stable scalability. Specifically, PSGNNs on average boost area under the curve (AUC) more than 14% for pairwise node classification and 18% for link prediction over the existing state-of-the-art position-aware methods. Our source code is publicly available at: https://github.com/ZhenyueQin/PSGNN.

Data augmentation has been widely used in machine learning for natural language processing and computer vision tasks to improve model performance. However, little research has studied data augmentation on graph neural networks, particularly using augmentation at both train- and test-time. Inspired by the success of augmentation in other domains, we have designed a method for social influence prediction using graph neural networks with train- and test-time augmentation, which can effectively generate multiple augmented graphs for social networks by utilising a variational graph autoencoder in both scenarios. We have evaluated the performance of our method on predicting user influence on multiple social network datasets. Our experimental results show that our end-to-end approach, which jointly trains a graph autoencoder and social influence behaviour classification network, can outperform state-of-the-art approaches, demonstrating the effectiveness of train-and test-time augmentation on graph neural networks for social influence prediction. We observe that this is particularly effective on smaller graphs.

Network modeling is a key enabler to achieve efficient network operation in future self-driving Software-Defined Networks. However, we still lack functional network models able to produce accurate predictions of Key Performance Indicators (KPI) such as delay, jitter or loss at limited cost. In this paper we propose RouteNet, a novel network model based on Graph Neural Network (GNN) that is able to understand the complex relationship between topology, routing, and input traffic to produce accurate estimates of the per-source/destination per-packet delay distribution and loss. RouteNet leverages the ability of GNNs to learn and model graph-structured information and as a result, our model is able to generalize over arbitrary topologies, routing schemes and traffic intensity. In our evaluation, we show that RouteNet is able to predict accurately the delay distribution (mean delay and jitter) and loss even in topologies, routing and traffic unseen in the training (worst case MRE=15.4%). Also, we present several use cases where we leverage the KPI predictions of our GNN model to achieve efficient routing optimization and network planning.

Analyses of the effectiveness of infectious disease control interventions often rely on dynamic transmission models to simulate intervention effects. We aim to understand how the choice of network or compartmental model can influence estimates of intervention effectiveness in the short and long term for an endemic disease with susceptible and infected states in which infection, once contracted, is lifelong. We consider 4 disease models with different permutations of socially connected network versus unstructured contact (mass-action mixing) model and heterogeneous versus homogeneous disease risk. The models have susceptible and infected populations calibrated to the same long-term equilibrium disease prevalence. We consider a simple intervention with varying levels of coverage and efficacy that reduces transmission probabilities. We measure the rate of prevalence decline over the first 365 d after the intervention, long-term equilibrium prevalence, and long-term effective reproduction ratio at equilibrium. Prevalence declined up to 10% faster in homogeneous risk models than heterogeneous risk models. When the disease was not eradicated, the long-term equilibrium disease prevalence was higher in mass-action mixing models than in network models by 40% or more. This difference in long-term equilibrium prevalence between network versus mass-action mixing models was greater than that of heterogeneous versus homogeneous risk models (less than 30%); network models tended to have higher effective reproduction ratios than mass-action mixing models for given combinations of intervention coverage and efficacy. For interventions with high efficacy and coverage, mass-action mixing models could provide a sufficient estimate of effectiveness, whereas for interventions with low efficacy and coverage, or interventions in which outcomes are measured over short time horizons, predictions from network and mass-action models diverge, highlighting the importance of sensitivity analyses on model structure. • We calibrate 4 models-socially connected network versus unstructured contact (mass-action mixing) model and heterogeneous versus homogeneous disease risk-to 10% preintervention disease prevalence.• We measure the short- and long-term intervention effectiveness of all models using the rate of prevalence decline, long-term equilibrium disease prevalence, and effective reproduction ratio.• Generally, in the short term, prevalence declined faster in the homogeneous risk models than in the heterogeneous risk models.• Generally, in the long term, equilibrium disease prevalence was higher in the mass-action mixing models than in the network models, and the effective reproduction ratio was higher in network models than in the mass-action mixing models.

As machine learning becomes more widely adopted across domains, it is critical that researchers and ML engineers think about the inherent biases in the data that may be perpetuated by the model. Recently, many studies have shown that such biases are also imbibed in Graph Neural Network (GNN) models if the input graph is biased, potentially to the disadvantage of underserved and underrepresented communities. In this work, we aim to mitigate the bias learned by GNNs by jointly optimizing two different loss functions: one for the task of link prediction and one for the task of demographic parity. We further implement three different techniques inspired by graph modification approaches: the Global Fairness Optimization (GFO), Constrained Fairness Optimization (CFO), and Fair Edge Weighting (FEW) models. These techniques mimic the effects of changing underlying graph structures within the GNN and offer a greater degree of interpretability over more integrated neural network methods. Our proposed models emulate microscopic or macroscopic edits to the input graph while training GNNs and learn node embeddings that are both accurate and fair under the context of link recommendations. We demonstrate the effectiveness of our approach on four real world datasets and show that we can improve the recommendation fairness by several factors at negligible cost to link prediction accuracy.

Network modeling is a critical component for building self-driving Software-Defined Networks, particularly to find optimal routing schemes that meet the goals set by administrators. However, existing modeling techniques do not meet the requirements to provide accurate estimations of relevant performance metrics such as delay and jitter. In this paper we propose a novel Graph Neural Network (GNN) model able to understand the complex relationship between topology, routing and input traffic to produce accurate estimates of the per-source/destination pair mean delay and jitter. GNN are tailored to learn and model information structured as graphs and as a result, our model is able to generalize over arbitrary topologies, routing schemes and variable traffic intensity. In the paper we show that our model provides accurate estimates of delay and jitter (worst case $R^2=0.86$) when testing against topologies, routing and traffic not seen during training. In addition, we present the potential of the model for network operation by presenting several use-cases that show its effective use in per-source/destination pair delay/jitter routing optimization and its generalization capabilities by reasoning in topologies and routing schemes not seen during training.