Feature Fusion‐Based Hybrid Quantum‐Classical Graph Residual Neural Network

Accurate classification of rice plant diseases is essential for early intervention and precision agriculture. However, real-world datasets often suffer from complex backgrounds, high-dimensional features, and severe class imbalances, which compromise classification performance. This study proposes an integrated framework combining image segmentation using U-Net, feature selection via Ant Colony Optimization (ACO), hybrid sampling to handle class imbalance, and final classification using a Convolutional Neural Network (CNN). Segmentation isolates disease-affected areas, ACO optimizes feature subsets, and hybrid sampling balances class distribution using undersampling and SMOTE. The proposed method was tested on four rice leaf disease datasets—Brown Spot, Leaf Blast, Leaf Blight, and Leaf Scald—exhibiting significant class imbalance. Experimental results show that the proposed approach outperforms baseline models (SegNet, PspNet, and E-Net) across multiple metrics: Accuracy, IoU, Precision, and Recall. This indicates the framework’s robustness and potential for real-world deployment in precision agriculture. Future work will focus on model compression and real-time implementation in IoT systems.

Class imbalance is an important problem in machine learning. With increases in available information and the growing use of Big Data sources to extract meaning from data, the challenges associated with class imbalance continue to influence research and shape business value. In this paper, we focus on using highly imbalanced Big Data from Medicare to detect provider claims fraud. We combine three Medicare parts and generate fraud labels using real-world excluded providers. The number of known fraudulent providers is very small, with 0.062% of the combined dataset being labeled as fraud, indicating severe class imbalance. To address class imbalance concerns, we provide experimental results incorporating six different data sampling methods (undersampling and oversampling) to create datasets for five class ratios (imbalanced to balanced), as well as using the full dataset (with no sampling). Three state-of-the-art machine learning models with Apache Spark are used to assess Medicare fraud detection performance across data sampling methods and class ratios. We demonstrate that data sampling, in particular random undersampling, presents good results across all learners, whereas oversampling provides no benefit versus models built using the full dataset.

Federated Learning (FL) is a data-decentralized learning paradigm that preserves data privacy by enabling model training without sharing raw data. However, FL suffers from performance degradation caused by class imbalance, a common form of statistical heterogeneity in non-independent and identically distributed environments. This paper proposes a modified FedAvg aggregation strategy that adjusts client weights using the ratio between the minimum and maximum class-wise average losses observed during training, reflecting the reliability of each local model. Unlike previous methods, it does not require prior knowledge of class distributions and dynamically adjusts aggregation weights based on client-wise reliability. Experiments on the CIFAR-10 dataset partitioned with a Dirichlet distribution demonstrate that the proposed method improves global model accuracy by 4–5% under severe class imbalance while maintaining comparable performance under more balanced conditions. Additionally, the method improves the stability of local updates for clients with extremely imbalanced data distributions. These findings demonstrate the effectiveness of the proposed weighting strategy in mitigating performance degradation caused by class imbalance in FL environments.

Class imbalance and the curse of minority hubs

Climate change poses significant challenges to agriculture, leading to increased crop damage owing to extreme weather conditions. Detecting and analyzing such damage is crucial for mitigating its effects on crop yield. This study proposes a novel autoencoder (AE)-based model, termed “Memory Ganomaly,” designed to detect and analyze weather-induced crop damage under conditions of significant class imbalance. The model integrates memory modules into the Ganomaly architecture, thereby enhancing its ability to identify anomalies by focusing on normal (undamaged) states. The proposed model was evaluated using apple and peach datasets, which included both damaged and undamaged images, and was compared with existing robust Convolutional neural network (CNN) models (ResNet-50, EfficientNet-B3, and ResNeXt-50) and AE models (Ganomaly and MemAE). Although these CNN models are not the latest technologies, they are still highly effective for image classification tasks and are deemed suitable for comparative analyses. The results showed that CNN and Transformer baselines achieved very high overall accuracy (94–98%) but completely failed to identify damaged samples, with precision and recall equal to zero under severe class imbalance. Few-shot learning partially alleviated this issue (up to 75.1% recall in the 20-shot setting for the apple dataset) but still lagged behind AE-based approaches in terms of accuracy and precision. In contrast, the proposed Memory Ganomaly delivered a more balanced performance across accuracy, precision, and recall (Apple: 80.32% accuracy, 79.4% precision, 79.1% recall; Peach: 81.06% accuracy, 83.23% precision, 80.3% recall), outperforming AE baselines in precision and recall while maintaining comparable accuracy. This study concludes that the Memory Ganomaly model offers a robust solution for detecting anomalies in agricultural datasets, where data imbalance is prevalent, and suggests its potential for broader applications in agricultural monitoring and beyond. While both Ganomaly and MemAE have shown promise in anomaly detection, they suffer from limitations—Ganomaly often lacks long-term pattern recall, and MemAE may miss contextual cues. Our proposed Memory Ganomaly integrates the strengths of both, leveraging contextual reconstruction with pattern recall to enhance detection of subtle weather-related anomalies under class imbalance.

Credit card fraud detection remains challenging due to extreme class imbalance, evolving fraud patterns, and asymmetric misclassification costs. This paper presents a deployment-oriented evaluation and decision-calibration framework for transaction-level fraud detection on the public creditcard.csv dataset, assessed under a strictly chronological train–validation–test protocol that mirrors real-world operation. Our contribution lies in evaluation design and cost-aware decision calibration rather than architectural novelty, focusing on how probabilistic model outputs are translated into operational decisions under explicit cost constraints. A standard feed-forward deep neural network (DNN) is trained on the numerical features using a class-weighted binary cross-entropy loss, with early stopping guided by validation AUC–PR. At deployment time, the decision threshold is selected on the validation window by minimizing an empirical cost function that penalizes false negatives more than false positives. On the held-out test set, the proposed pipeline achieves a ROC-AUC of 0.9489 and a PR-AUC of 0.7813. We show that decision policy choice strongly affects operational outcomes: naive thresholding yields excessive false alarms, whereas validation-based cost-sensitive calibration substantially reduces expected loss. Under C_FN = 10 and C_FP = 1, the cost-optimal threshold yields an expected test cost of 190. Comparisons with logistic regression, random forest, and XGBoost under identical preprocessing, temporal splitting, and decision calibration show that tree-based ensembles remain highly competitive, while the evaluated DNN achieves comparable cost and precision–recall performance. Overall, the results highlight the importance of combining chronological evaluation with explicit cost-sensitive thresholding for practical fraud detection under severe class imbalance.

Stance detection has gained increasing interest from the research community due to its importance for fake news detection. The goal of stance detection is to categorize an overall position of a subject towards an object into one of the four classes: agree, disagree, discuss, and unrelated. One of the major problems faced by current machine learning models used for stance detection is caused by a severe class imbalance among these classes. Hence, most models fail to correctly classify instances that fall into minority classes. In this paper, we address this problem by proposing a hierarchical representation of these classes, which combines the agree, disagree, and discuss classes under a new related class. Further, we propose a two-layer neural network that learns from this hierarchical representation and controls the error propagation between the two layers using the Maximum Mean Discrepancy regularizer. Compared with conventional four-way classifiers, this model has two advantages: (1) the hierarchical architecture mitigates the class imbalance problem; (2) the regularization makes the model to better discern between the related and unrelated stances. An extensive experimentation demonstrates state-of-the-art accuracy performance of the proposed model for stance detection.

Data bias in geohazard artificial intelligence (AI) systems, particularly class distribution imbalances, critically undermines the reliability of landslide detection models. While active learning (AL) offers promise for mitigating annotation costs and addressing data biases, the interplay between landslide class proportions and AL efficiency remains poorly quantified; additionally, self-optimizing mechanisms to adaptively manage class imbalances are underexplored. This study bridges these gaps by rigorously evaluating how landslide-to-non-landslide ratios (1:1, 1:12, and 1:30) influence the effectiveness of a widely used AL strategy—margin sampling. Leveraging open-source landslide inventories, we benchmark margin sampling against random sampling using the area under the receiver operating characteristic curve (AUROC) and partial AUROC while analyzing spatial detection accuracy through classification maps. The results reveal that margin sampling significantly outperforms random sampling under severe class imbalances (1:30), achieving 12–18% higher AUROC scores and reducing false negatives in critical landslide zones. In balanced scenarios (1:1), both strategies yield comparable numerical metrics; however, margin sampling produces spatially coherent detections with fewer fragmented errors. These findings indicate that regardless of the landslide proportion, AL enhances the generalizability of landslide detection models in terms of predictive accuracy and spatial consistency. This work also provides actionable guidelines for deploying adaptive AI systems in data-scarce, imbalance-prone environments.

• Multi-sine AC signals enable rapid, precise real-time PEMFC fault diagnosis. • SMOTE-Tomek balances and clears noisy data, sharply defining rare fault conditions. • Dual-attention WGAN synthesizes realistic PEMFC fault signals with high fidelity. • Achieves 99.67 % accuracy, surpassing state-of-the-art by 14 % at high imbalance. • Robust diagnosis and high accuracy using limited and imbalanced fault data. Proton exchange membrane fuel cell (PEMFC) faults, especially dehydration and flooding, cause distinct changes in electrochemical behavior. Consequently, real-time monitoring is essential for early and accurate diagnosis. However, acquiring real-world fault data is challenging, and the rarity of such faults results in severe class imbalance. This imbalance limits the performance, reliability, and practical applicability of conventional diagnostic methods. To address these limitations, this study proposes a unified diagnostic framework that integrates multi-sine AC voltage response, boundary-aware resampling, and attention-guided generative modeling. The key innovations of the proposed approach include: (1) Enhanced fault separability through the first application of multi-sine AC voltage response under data imbalance, enabling real-time extraction of critical electrochemical spectral features for early-stage diagnosis; (2) Improved data balance and clearer class boundaries using synthetic minority oversampling with Tomek links, which oversamples minority classes and removes borderline samples; (3) Realistic minority class synthesis using a dual attention Wasserstein generative adversarial networks, where channel attention focuses on diagnostically relevant spectral features and temporal attention models the dynamic evolution of PEMFC electrochemical behavior, ensuring high-quality, diagnostically informative synthetic fault data. The integrated framework achieves 99.67 % overall diagnostic accuracy and, under an extreme 1:200 class imbalance, outperforms state-of-the-art methods by 14 %. This approach enables rapid, data-efficient PEMFC fault diagnosis, strengthening fault management and advancing the performance of energy systems.

Bank telemarketing campaigns often experience low subscription rates due to customer heterogeneity and severe class imbalance, which pose challenges for reliable predictive modeling. This study investigates a data-driven approach that integrates synthetic minority oversampling and cost-sensitive learning to improve the prediction of telemarketing outcomes. Experiments are conducted using the Portuguese Bank Marketing dataset, comprising 41,188 instances with a positive response rate of 11.3%. Eight machine learning models are evaluated under a unified preprocessing pipeline and five-fold stratified cross-validation, including Logistic Regression, Decision Tree, Random Forest, and Ensemble methods. The results show that Ensemble models, particularly CatBoost, XGBoost, and LightGBM, achieve improved performance compared with traditional baselines, with notable gains in minority-class recall and overall discrimination ability. The best-performing model attains an F1-score of 0.540, a recall of 0.812 for the positive class, and a ROC–AUC of 0.908. To enhance interpretability, SHAP-based analysis is applied to quantify feature contributions, identifying campaign duration, previous contact outcomes, and selected macroeconomic indicators as key predictors. These findings indicate that combining resampling strategies with cost-sensitive optimization provides a robust and transparent approach for learning from imbalanced telemarketing data, thereby supporting reproducible and data-driven financial decision-making by explicitly addressing difficulty in minority-class identification under imbalance and class imbalance under cross-entropy training in imbalanced banking data.

Drug-drug interactions (DDIs) are a significant focus in biomedical research and clinical practice due to their potential to compromise treatment outcomes or cause adverse effects. While deep learning approaches have advanced DDI extraction, challenges such as severe class imbalance and the complexity of biomedical relationships persist. This study introduces BioFocal-DDI, a framework combining BioGPT for data augmentation, BioBERT and BiLSTM for contextual and sequential feature extraction, and Relational Graph Convolutional Networks (ReGCN) for relational modeling. To address class imbalance, a Focal Loss-based Attention mechanism is employed to enhance learning on underrepresented and challenging instances. Evaluated on the DDI Extraction 2013 dataset, BioFocal-DDI achieves a precision of 86.75%, recall of 86.53%, and an F1 Score of 86.64%. These results suggest that the proposed method is effective in improving DDI extraction.

Deep learning based diagnostic AI systems based on medical images are starting to provide similar performance as human experts. However, these data-hungry complex systems are inherently black boxes and therefore slow to be adopted for high-risk applications like healthcare. This problem of lack of transparency is exacerbated in the case of recent large foundation models, which are trained in a self-supervised manner on millions of data points to provide robust generalisation across a range of downstream tasks. The embeddings generated from them happen through a process that is not interpretable, and hence not easily trustable for clinical applications. To address this timely issue, we deploy conformal analysis to quantify the predictive uncertainty of a vision transformer (ViT)-based foundation model across patient demographics with respect to sex, age, and ethnicity for the task of skin lesion classification using several public benchmark datasets. The significant advantage of this method is that conformal analysis is method independent, and it not only provides a coverage guarantee at the population level but also provides an uncertainty score for each individual. This is used to demonstrate the effectiveness of utilizing these embeddings for specialized tasks like diagnostic classification, meanwhile reducing computational costs. Secondly, the public benchmark datasets we used had severe class imbalance in terms of the number of samples in different classes. We used a model-agnostic dynamic F1-score-based sampling during model training, which helped to stabilize the class imbalance. We investigate the effects on uncertainty quantification (UQ) with or without this bias mitigation step. Thus, our results show how this can be used as a fairness metric to evaluate the robustness of the feature embeddings of the foundation model (Google DermFoundation), advancing the trustworthiness and fairness of clinical AI.

Purpose: High-reliability systems rarely fail, resulting in extreme fault-data scarcity and severe class imbalance. These limitations hinder the application of machine learning models for Remaining Useful Life (RUL) prediction and fault diagnosis. This study proposes a hybrid data augmentation framework to overcome these challenges.Methods: A CTGAN-based hybrid augmentation approach was developed by combining environmental sensitivity analysis (XGBoost-SHAP), prototype clustering (K-means), and multi-stage post-processing (ProtoMix, weighted noise injection, Soft Capping, and CDF matching). The framework generates physically consistent and statistically realistic synthetic fault data.Results: Under an extreme stress test with 37,982 normal samples and only 15 fault samples, the Hybrid GAN outperformed Baseline, SMOTE, and Basic GAN. It achieved the highest accuracy (0.942), improved F1-score (0.905), precision (0.964), and reduced false positives by nearly 50% compared with the Basic GAN.Conclusion: The hybrid framework effectively mitigated class imbalance and delivered robust predictive performance under challenging conditions, demonstrating strong potential for predictive maintenance in high-reliability systems.

Purpose To detect academic misconduct in students' assessment score trajectories under severe class imbalance. The paper compares tabular learners, gradient-boosting models, and sequence-aware deep networks, and proposes a precision–recall–centric evaluation and deployment protocol (calibration, threshold selection and Recall@Top-k%) tailored to rare-event screening in educational settings. Design/methodology/approach A cohort of 1,527 students (2021–2024) is modeled using ten algorithms: LR, DT, RF, SVM, MLP, XGBoost, CatBoost, LightGBM, GRU-RNN and 1D-CNN. Features encode sequential score dynamics and metadata. Models are tuned via cross-validation; probabilities are calibrated (Platt/Isotonic); operating thresholds are chosen on validation to maximize minority-class F1 or a cost-sensitive utility. Performance is assessed on a hold-out test set with PR-AUC (headline), F1(+), Recall@Top-k%, ROC-AUC, calibration curves, Brier, and bootstrap CIs. Findings Sequence-aware models dominate: GRU-RNN and 1D-CNN achieve ROC-AUC ˜0.97–0.98 and the highest F1(+) and Recall@Top-5%. Tabular/boosting baselines show ˜0.90 accuracy yet miss most positives at the default 0.5 threshold, highlighting the necessity of calibration and threshold optimization. With PR-centric selection and tuned operating points, deep temporal models yield strong screening utility for limited human review budgets. Research limitations/implications Labels reflect suspected–not adjudicated–cheating, introducing noise. The single-institution cohort may limit external validity; temporal shift across semesters can degrade performance. Future work should include multi-site evaluation, collusion/graph modeling, semi-/weak-supervision for noisy labels and governance topics (fairness audits, drift monitoring and uncertainty reporting). Practical implications This study highlights how educational institutions can leverage machine learning for early detection of academic dishonesty based on historical performance data. CNN and RNN models are promising tools for identifying anomalous learning patterns. However, practical deployment requires preprocessing techniques to manage class imbalance and threshold optimization to reduce false negatives. The findings provide a roadmap for building automated cheating detection systems in both online and traditional assessment environments. Social implications By improving the ability to detect cheating, this research contributes to fairer academic environments, upholding educational integrity and credibility of credentials. However, ethical considerations must be taken into account to avoid false accusations and ensure student rights. Human-in-the-loop systems are crucial for verifying algorithmic predictions before disciplinary action, thereby fostering transparency and accountability in automated decision-making processes. Originality/value The study unifies a minority-focused evaluation protocol with a comprehensive comparison of tabular, boosting, and sequence-aware models for cheating detection from score trajectories. It demonstrates the decisive value of temporal representation learning and provides a reproducible pipeline and operational metrics that align model performance with real investigative workflows.

<title>Abstract</title> Credit card fraud detection remains a critical yet challenging task due to the extreme class imbalance inherent in transaction datasets, where fraudulent activities constitute only a small fraction of the total records. To address this imbalance and enhance the detection of rare fraud instances, this study proposes a novel hybrid framework that integrates density-based clustering for data augmentation with an ensemble classification strategy optimized for high recall. In the preprocessing stage, the proposed method utilizes Density-Based Spatial Clustering of Applications with Noise (DBSCAN) to identify minority-class clusters and synthetically augment the fraud class. This step aims to preserve the intrinsic structure of fraudulent patterns while increasing their representation in the training set. Subsequently, an ensemble model comprising Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Machine (SVM) classifiers is constructed. Final predictions are generated using a Disjunctive Voting Ensemble (DVE) strategy, wherein a transaction is classified as fraudulent if any of the base classifiers predicts it as such. This permissive voting mechanism prioritizes recall, thereby minimizing the risk of undetected fraudulent transactions. Extensive experiments conducted on three publicly available credit card fraud imbalanced datasets containing transaction records from European credit cardholders in 2023, providing a realistic fraud detection scenario. Computational efficiency in both training and testing is achieved using the Anaconda Navigator (Spider-Python 3.12) environment. The comparative analysis shows that the proposed DBSCAN-augmented DVE framework delivers notable improvements over traditional ensemble approaches and single-model baselines, particularly in recall and F1-score, while preserving consistently high precision. This approach proves to be both robust and interpretable, making it highly applicable to real-world fraud detection scenarios characterized by severe class imbalance. The results are especially compelling, achieving recall and F1-scores as high as 99.5% and 99.8% respectively, and consistently maintaining a perfect accuracy and precision of 100% across all imbalanced datasets. This study highlights the effectiveness of hybrid ensemble approaches in combating credit card fraud. The findings lay the groundwork for developing more resilient and adaptive fraud detection systems, which are crucial in countering the evolving tactics of fraudsters. The proposed model marks a significant advancement in securing financial transactions and mitigating risks in an increasingly digital economy.