In the digital era, effective Transaction Fraud Detection (TFD) is essential to ensuring financial security. The considerable class imbalance, with legitimate transactions vastly outnumbering fraudulent ones, presents a significant challenge for TFD models to accurately identify fraudulent patterns. While existing sample-balancing strategies address class imbalance effectively in many contexts, they often fall short in TFD due to fraudsters’ sophisticated concealment tactics, which lead to pronounced behavioral overlap between fraudulent and legitimate transactions. In this paper, we introduce a novel Generative Adversarial Network-based Hybrid Sampling method (GANHS) to effectively address the class imbalance issue. GANHS employs a dual-discriminator generative adversarial network to generate synthetic samples that accurately reflect the characteristics of fraudulent activity, while an adaptive neighborhood-based undersampling technique refines these samples to minimize overlap with legitimate ones. This hybrid approach not only enhances the model’s ability to learn fraud patterns by generating high-quality samples but also improves its resilience against highly concealed fraudulent activities. Experiments on real-world and public datasets demonstrate that GANHS outperforms its competitive peers, with gains of 0.5%–8.7% in average$F_{1}$-Score and 1.0%–7.0% in G-mean, highlighting its strong potential for improving the reliability and effectiveness of TFD systems in complex, high-risk financial scenarios.

Most of current anomaly detection models assume that the normal pattern remains the same all the time. However, the normal patterns of web services can change dramatically and frequently over time. The model trained on old-distribution data becomes outdated and ineffective after such changes. Retraining the whole model whenever the pattern is changed is computationally expensive. Further, at the beginning of normal pattern changes, there is not enough observation data from the new distribution. Retraining a large neural network model with limited data is vulnerable to overfitting. Thus, we propose a Light Anti-overfitting Retraining Approach (LARA) based on deep variational auto-encoders for time series anomaly detection. In LARA we make the following three major contributions: 1) the retraining process is designed as a convex problem such that overfitting is prevented and the retraining process can converge fast; 2) a novel ruminate block is introduced, which can leverage the historical data without the need to store them; 3) we mathematically and experimentally prove that when fine-tuning the latent vector and reconstructed data, the linear formations can achieve the least adjusting errors between the ground truths and the fine-tuned ones. Moreover, we have performed many experiments to verify that retraining LARA with even a limited amount of data from new distribution can achieve competitive performance in comparison with the state-of-the-art anomaly detection models trained with sufficient data. Besides, we verify its light computational overhead.

Feature selection can be used to solve imbalanced classification problems encountered in big data projects. There often exist multiple feature subsets achieving the same accuracy. These subsets tend to exhibit different acquisition difficulty and reliability, thus offering decision-makers with multiple choices if they can be well-identified. This work formulates feature selection as a Multimodal Multiobjective Problem (MMOP), where a point on Pareto front in objective space has multiple equivalent feature subsets in decision space. To seek more equivalent feature subsets, this work proposes a new multiobjective fireworks algorithm. It extends a latest single-objective fireworks algorithm to a multiobjective version such that it becomes suitable for solving MMOP. An adaptive strategy and special archive guidance are newly designed to improve its performance. A weighted extreme learning machine is chosen to classify datasets and return classification accuracy due to its fast learning speed. Experimental results show that the proposed algorithm outperforms its compared ones on 15 imbalanced classification datasets including 5 low-dimensional, 5 high-dimensional feature selection problems and 5 large-scale problems with larger imbalanced ratio, and its runtime is the least among them. Also, fault diagnosis in self-organizing cellular networks, as an important imbalance classification problem, is performed by the proposed algorithm and the results show that it can perform fault diagnosis well.

A fast non-negative latent factor (FNLF) model for a high-dimensional and sparse (HiDS) matrix adopts a Single Latent Factor-dependent, Non-negative, Multiplicative and Momentum-incorporated Update (SLF-NM2U) algorithm, which enables its fast convergence. It is crucial to achieve a rigorously theoretical proof regarding its fast convergence, which has not been provided in prior research. Aiming at addressing this critical issue, this work theoretically proves that with an appropriately chosen momentum coefficient, SLF-NM2U enables the fast convergence of an FNLF model in both continuous and discrete time cases. Empirical analysis of HiDS matrices generated by representative industrial applications provides empirical evidences for the theoretical proof. Hence, this study represents an important milestone in the field of HiDS matrix analysis.

Graphs are widely used to model various practical applications. In recent years, graph convolution networks (GCNs) have attracted increasing attention due to the extension of convolution operation from traditional grid data to graph one. However, the representation ability of current GCNs is undoubtedly limited because existing work fails to consider feature interactions. Toward this end, we propose a Dual Feature Interaction-based GCN. Specifically, it models feature interaction in the aspects of 1) node features where we use Newton's identity to extract different-order cross features implicit in the original features and design an attention mechanism to fuse them; and 2) graph convolution where we capture the pairwise interactions among nodes in the neighborhood to expand a weighted sum operation. We evaluate the proposed model with graph data from different fields, and the experimental results on semi-supervised node classification and link prediction demonstrate the effectiveness of the proposed GCN. The data and source codes of this work are available athttps://github.com/ZZY-GraphMiningLab/DFI-GCN.

Fuzzy Petri nets (FPNs) are a useful instrument for modelling expert systems to conduct knowledge representation and reasoning. Many studies have been carried out for improving the performance of FPNs in terms of their accurate representation of knowledge and power of approximate reasoning. Nevertheless, the current representation methods with FPNs are unable to handle the uncertain linguistic knowledge given by domain experts and the reliability of their judgments. In addition, the existing reasoning algorithms have no way to capture the interrelationship of the propositions with the same output transition. Therefore, we present a new type of FPNs, called 2-dimensional uncertain linguistic Petri nets (2DULPNs). The 2-dimensional uncertain linguistic variables (2DULVs) and Choquet integral are combined for knowledge representation and reasoning for the first time. The truth degrees of propositions, thresholds and certainty values of linguistic production rules are denoted as 2DULVs. Some new aggregated operators based on Choquet integral are proposed and used in the approximate reasoning to capture the interactions among antecedent propositions. Finally, an equipment fault diagnosis example is provided to illustrate the correctness and effectiveness of the proposed 2DULPN model.

Parkinson’s disease is a progressive nervous system disorder afflicting millions of patients. Among its motor symptoms, bradykinesia is one of the cardinal manifestations. Experienced doctors are required for the clinical diagnosis of bradykinesia, but sometimes they also miss subtle changes, especially in early stages of such disease. Therefore, developing auxiliary diagnostic methods that can automatically detect bradykinesia has received more and more attention. In this article, we employ a two-stage framework for bradykinesia recognition based on the video of patient movement. First, convolution neural networks are trained to localize keypoints in each video frame. These time-varying coordinates form motion trajectories that represent the whole movement. From the trajectory, we then propose novel measurements, namely stability , completeness , and self-similarity , to quantify different motor behaviors. We also propose a periodic motion model called PMNet . An encoder--decoder structure is applied to learn a low dimensional representation of a motion process. The compressed motion process and quantified motor behaviors are combined as inputs to a fully-connected neural network. Different from the traditional means, our solution extends the application scenario outside the hospital and can be easily transplanted to conduct similar tasks. A commonly used clinical assessment is served as a case study. Experimental results based on real-world data validate the effectiveness of our approach for bradykinesia recognition.