Search Results (581)

Receiver operating characteristic (ROC) surfaces provide a natural extension of ROC curves to three-class diagnostic problems. A key summary index is the volume under the surface (VUS), representing the probability that a randomly chosen observation from each of the three ordered groups is correctly classified. A parametric estimation of VUS typically assumes trinormality of the class distributions. However, a formal method for the verification of this composite assumption has not appeared in the literature. Our approach generalizes the two-class AUC-based GOF test of Zou et al. to the three-class setting by exploiting the parallel structure between empirical and trinormal VUS estimators. We propose a global goodness-of-fit (GOF) test for trinormal ROC models based on the difference between empirical and trinormal parametric estimates of the VUS. To improve stability, a probit transformation is applied and a bootstrap procedure is used to estimate the variance of the difference. The resulting test provides a formal diagnostic for assessing the adequacy of trinormal ROC modeling. Simulation studies illustrate the robustness of the assumption via the empirical size and power of the test under various distributional settings, including skewed and multimodal alternatives. The method’s application to COVID-19 antibody level data demonstrates the practical utility of it. Our findings suggest that the proposed GOF test is simple to implement, computationally feasible for moderate sample sizes, and a useful complement to existing ROC surface methodology. Full article

Accurate automatic sleep stage classification from single-channel EEG remains challenging due to the need for effective extraction of multiscale neurophysiological features and modeling of long-range temporal dependencies. This study aims to address these limitations by developing an efficient and compact deep learning architecture tailored for wearable and edge device applications. We propose MultiScaleSleepNet, a hybrid convolutional neural network–bidirectional long short-term memory–transformer architecture that extracts multiscale temporal and spectral features through parallel convolutional branches, followed by sequential modeling using a BiLSTM memory network and transformer-based attention mechanisms. The model obtained an accuracy, macro-averaged F1 score, and kappa coefficient of 88.6%, 0.833, and 0.84 on the Sleep-EDF dataset; 85.6%, 0.811, and 0.80 on the Sleep-EDF Expanded dataset; and 84.6%, 0.745, and 0.79 on the SHHS dataset. Ablation studies indicate that attention mechanisms and spectral fusion consistently improve performance, with the most notable gains observed for stages N1, N3, and rapid eye movement. MultiScaleSleepNet demonstrates competitive performance across multiple benchmark datasets while maintaining a compact size of 1.9 million parameters, suggesting robustness to variations in dataset size and class distribution. The study supports the feasibility of real-time, accurate sleep staging from single-channel EEG using parameter-efficient deep models suitable for portable systems. Full article

In image classification tasks, models may occasionally produce incorrect predictions, which can lead to severe consequences in safety-critical applications. For instance, if a model mistakenly classifies a red traffic light as green, it could result in a traffic accident. Therefore, it is essential to assess the confidence level associated with each prediction. Predictions accompanied by high confidence scores are generally more reliable and can serve as a basis for informed decision-making. To address this, the present paper extends the block-scaling approach—originally developed for estimating classifier accuracy on unlabeled datasets—to compute confidence scores for individual samples in image classification. The proposed method, termed block masking confidence (BMC), applies a sliding mask filled with random noise to occlude localized regions of the input image. Each masked variant is classified, and predictions are aggregated across all variants. The final class is selected via majority voting, and a confidence score is derived based on prediction consistency. To evaluate the effectiveness of BMC, we conducted experiments comparing it against Monte Carlo (MC) dropout and a vanilla baseline across image datasets of varying sizes and distortion levels. While BMC does not consistently outperform the baselines under standard (in-distribution) conditions, it shows clear advantages on distorted and out-of-distribution (OOD) samples. Specifically, on the level-3 distorted iNaturalist 2018 dataset, BMC achieves a median expected calibration error (ECE) of 0.135, compared to 0.345 for MC dropout and 0.264 for the vanilla approach. On the level-3 distorted Places365 dataset, BMC yields an ECE of 0.173, outperforming MC dropout (0.290) and vanilla (0.201). For OOD samples in Places365, BMC achieves a peak entropy of 1.43, higher than the 1.06 observed for both MC dropout and vanilla. Furthermore, combining BMC with MC dropout leads to additional improvements. On distorted Places365, the median ECE is reduced to 0.151, and the peak entropy for OOD samples increases to 1.73. Overall, the proposed BMC method offers a promising framework for uncertainty quantification in image classification, particularly under challenging or distribution-shifted conditions. Full article

Automatic detection of clue cells is crucial for rapid diagnosis of bacterial vaginosis (BV), but existing algorithms suffer from low sensitivity. This is because clue cells are highly similar to normal epithelial cells in terms of macroscopic size and shape. The key difference between clue cells and normal epithelial cells lies in the surface texture and edge morphology. To address this specific problem, we propose an clue cell detection algorithm named CA-YOLO. The contributions of our approach lie in two synergistic and custom-designed feature extraction modules: the context-aware module (CAM) extracts and captures bacterial distribution patterns on the surface of clue cells; and the shuffle global attention mechanism (SGAM) enhances cell edge features and suppresses irrelevant information. In addition, we integrate focal loss into the classification loss to alleviate the severe class imbalance problem inherent in clinical samples. Experimental results show that the proposed CA-YOLO achieves a sensitivity of 0.778, which is 9.2% higher than the baseline model, making the automated BV detection more reliable and feasible. Full article

This paper introduces a new class of probability distributions, termed the generated log exponentiated polynomial (GLEP) family, designed to enhance flexibility in modeling complex real financial data. The proposed family is constructed through a novel cumulative distribution function that combines logarithmic and exponentiated polynomial structures, allowing for rich distributional shapes and tail behaviors. We present comprehensive mathematical properties, including useful series expansions for the density, cumulative, and quantile functions, which facilitate the derivation of moments, generating functions, and order statistics. Characterization results based on the reverse hazard function and conditional expectations are established. The model parameters are estimated using various frequentist methods, including Maximum Likelihood Estimation (MLE), Cramer–von Mises (CVM), Anderson–Darling (ADE), Right Tail Anderson–Darling (RTADE), and Left Tail Anderson–Darling (LEADE), with a comparative simulation study assessing their performance. Risk analysis is conducted using actuarial key risk indicators (KRIs) such as Value-at-Risk (VaR), Tail Value-at-Risk (TVaR), Tail Variance (TV), Tail Mean Variance (TMV), and excess function (EL), demonstrating the model’s applicability in financial and insurance contexts. The practical utility of the GLEP family is illustrated through applications to real and simulated datasets, including house price dynamics and insurance claim sizes. Peaks Over Random Threshold Value-at-Risk (PORT-VaR) analysis is applied to U.K. motor insurance claims and U.S. house prices datasets. Some recommendations are provided. Finally, a comparative study is presented to prove the superiority of the new family. Full article

We propose a life-expectancy pay-off function (LEP) for determining optimal cancer treatment within a control theory framework. The LEP averages life expectancy over all future outcomes, outcomes that are determined by key events during therapy such as tumour elimination (cure) and patient death (including treatment related mortality). We analyse this optimisation problem for tumours treated with chemotherapy using tumour growth models based on ordinary differential equations. To incorporate tumour elimination we draw on branching processes to compute the probability distribution of tumour population extinction. To demonstrate the approach, we apply the LEP framework to simplified one-compartment models of tumour growth that include three possible outcomes: cure, relapse, or death during treatment. Using Pontryagin’s maximum principle (PMP) we show that the best treatment strategies fall into three categories: (i) continuous treatment at the maximum tolerated dose (MTD), (ii) no treatment, or (iii) treat-and-stop therapy, where the drug is given at the MTD and then halted before the treatment (time) horizon. Optimal treatment strategies are independent of the time horizon unless the time horizon is too short to accommodate the most effective (treat-and-stop) therapy. For sufficiently long horizons, the optimal solution is either no treatment (when treatment yields no benefit) or treat-and-stop. Patients, thus, split into an untreatable class and a treatable class, with patient demographics, tumour size, tumour response, and drug toxicity determining whether a patient benefits from treatment. The LEP is in principle parametrisable from data, requiring estimation of the rates of each event and the associated life expectancy under that event. This makes the approach suitable for personalising cancer therapy based on tumour characteristics and patient-specific risk profiles. Full article

Colorectal cancer (CRC) is driven by a complex spectrum of somatic mutations and structural variants that contribute to tumor heterogeneity and therapy resistance. In this study, we performed a comparative analysis of short-read Illumina and long-read Nanopore sequencing technologies across multiple CRC sample groups, encompassing diverse tissue morphologies. Our evaluation included general base-level metrics—such as nucleotide ratios, sequence match rates, and coverage—as well as variant calling performance, including variant allele frequency (VAF) distributions and pathogenic mutation detection rates. Focusing on clinically relevant genes (KRAS, BRAF, TP53, APC, PIK3CA, and others), we characterized platform-specific detection profiles and completed the ground truth validation of somatic KRAS and BRAF mutations. Structural variant (SV) analysis revealed Nanopore’s enhanced ability to resolve large and complex rearrangements, with consistently high precision across SV types, though recall varied by variant class and size. To enable direct comparison with the Illumina exome panel, we applied an exonic position reference file. To assess the impact of depth and PCR amplification, we completed an additional high-coverage Nanopore sequencing run. This analysis confirmed that PCR-free protocols preserve methylation signals more accurately, reinforcing Nanopore’s utility for integrated genomic and epigenomic profiling. Together, these findings underscore the complementary strengths of short- and long-read sequencing platforms in high-resolution cancer genomics, and we highlight the importance of coverage normalization, epigenetic fidelity, and rigorous benchmarking in variant discovery. Full article

Background: Class B carbapenemases confer high-level resistance to carbapenems in Klebsiella pneumoniae. In Brazil, data on the national burden and geographic distribution of these genes among clinical K. pneumoniae isolates are sporadic. We performed a systematic review and meta-analysis to estimate the prevalence of MβL genes in Brazilian clinical K. pneumoniae. Methods: We searched SciELO, PubMed, ScienceDirect and LILACS for original studies published between 2006 and 2024 reporting molecular detection of MβL in clinical K. pneumoniae isolates from Brazil. Articles were independently screened, along with the extracted data and appraised study quality using Joanna Briggs Institute checklists. A random-effects meta-analysis estimated the pooled prevalence of MβL producers and assessed heterogeneity and publication bias. Results: Fifteen studies including 3.533 clinical K. pneumoniae isolates met inclusion criteria. Overall, 402 isolates (11.4%) harbored MβL genes, yielding a pooled prevalence of 44.6%. Subgroup analysis demonstrated highest prevalence in the Southeast. bla_NDM was the dominant variant (present in 14/15 studies), with bla_VIM and bla_IMP rarely detected. Meta-regression revealed an inverse association between sample size and reported prevalence, and no significant publication bias was observed. Conclusions: MβLs, particularly NDM, are widespread in Brazilian clinical K. pneumoniae but show marked regional heterogeneity driven by differences in study design, laboratory capacity, and outbreak dynamics. Urgent expansion of standardized and multicenter molecular surveillance, including allele-specific detection, and strengthened laboratory infrastructure are needed and may inform targeted infection-control and antimicrobial-stewardship interventions. Full article

We propose a Transformer-based data augmentation framework with a time-series dual-stream architecture to address performance degradation in encrypted network traffic classification caused by class imbalance between attack and benign traffic. The proposed framework independently processes the complete flow’s sequential packet information and statistical characteristics by extracting and normalizing a local channel (comprising packet size, inter-arrival time, and direction) and a set of six global flow-level statistical features. These are used to generate a fixed-length multivariate sequence and an auxiliary vector. The sequence and vector are then fed into an encoder-only Transformer that integrates learnable positional embeddings with a FiLM + context token-based injection mechanism, enabling complementary representation of sequential patterns and global statistical distributions. Large-scale experiments demonstrate that the proposed method reduces reconstruction RMSE and additional feature restoration MSE by over 50%, while improving accuracy, F1-Score, and AUC by 5–7%p compared to classification on the original imbalanced datasets. Furthermore, the augmentation process achieves practical levels of processing time and memory overhead. These results show that the proposed approach effectively mitigates class imbalance in encrypted traffic classification and offers a promising pathway to achieving more robust model generalization in real-world deployment scenarios. Full article

Compressive Learning (CL) is an emerging paradigm that allows machine learning models to perform inference directly from compressed measurements, significantly reducing sensing and computational costs. While existing CL approaches have achieved competitive accuracy compared to traditional image-domain methods, they typically rely on reconstruction to address information loss and often neglect uncertainty arising from ambiguous or insufficient data. In this work, we propose MCS-TCL, a novel and trustworthy CL framework based on Multi-functional Compressive Sensing Sampling. Our approach unifies sampling, compression, and feature extraction into a single operation by leveraging the compatibility between compressive sensing and convolutional feature learning. This joint design enables efficient signal acquisition while preserving discriminative information, leading to feature representations that remain robust across varying sampling ratios. To enhance the model’s reliability, we incorporate evidential deep learning (EDL) during training. EDL estimates the distribution of evidence over output classes, enabling the model to quantify predictive uncertainty and assign higher confidence to well-supported predictions. Extensive experiments on image classification tasks show that MCS-TCL outperforms existing CL methods, achieving state-of-the-art accuracy at a low sampling rate of 6%. Additionally, our framework reduces model size by 85.76% while providing meaningful uncertainty estimates, demonstrating its effectiveness in resource-constrained learning scenarios. Full article

The survival status of Lumnitzera littorea is near threatened globally and critically endangered in China. Clarifying its global distribution pattern and its changing trends under different future climate models is of great significance for the protection and restoration of its endangered status. To build a model for this purpose, this study selected 73 actual distribution points of Lumnitzera littorea worldwide, combined with 12 environmental factors, and simulated its potential suitable habitats in six periods: the Last Interglacial (130,000–115,000 years ago), the Last Glacial Maximum (27,000–19,000 years ago), the Mid-Holocene (6000 years ago), the present (1970–2000), and the future 2050s (2041–2060) and 2070s (2061–2080). The results show that the optimal model parameter combination is the regularization multiplier RM = 4.0 and the feature combination FC (Feature class) = L (Linear) + Q (Quadratic) + P (Product). The MaxEnt model has a low omission rate and a more concise model structure. The AUC values in each period are between 0.981 and 0.985, indicating relatively high prediction accuracy. Min temperature of the coldest month, mean diurnal range, clay content, precipitation of the warmest quarter, and elevation are the dominant environmental factors affecting its distribution. The environmental conditions for min temperature of the coldest month at ≥19.6 °C, mean diurnal range at <7.66 °C, clay content at 34.14%, precipitation of the warmest quarter at ≥570.04 mm, and elevation at >1.39 m are conducive to Lumnitzera littorea’s survival and distribution. The global potential distribution areas are located along coasts. Starting from the paleoclimate, the plant’s distribution has gradually expanded, and its adaptability has gradually improved. In China, the range of potential highly suitable habitats is relatively narrow. Hainan Island is the core potential habitat, but there are fragmented areas in regions such as Guangdong, Guangxi, and Taiwan. The modern centroid of Lumnitzera littorea is located at (109.81° E, 2.56° N), and it will shift to (108.44° E, 3.22° N) in the later stage of the high-emission scenario (2070s (SSP585)). Under global warming trends, it has a tendency to migrate to higher latitudes. The development of the aquaculture industry and human deforestation has damaged the habitats of Lumnitzera littorea, and its population size has been sharply and continuously decreasing. The breeding and renewal system has collapsed, seed abortion and seedling establishment failure are common, and genetic variation is too scarce. This may indicate why Lumnitzera littorea is near threatened globally and critically endangered in China. Therefore, the protection and restoration strategies we propose are as follows: strengthen the legislative guarantee and law enforcement supervision of the native distribution areas of Lumnitzera littorea, expanding its population size outside the native environment, and explore measures to improve its seed germination rate, systematically collecting and introducing foreign germplasm resources to increase its genetic diversity. Full article

Bioactive peptides have significant potential for applications in pharmaceuticals, the food industry, and cosmetics due to their wide spectrum of biological activities. However, their pronounced structural and functional heterogeneity complicates the classification and prediction of biological activity. This study uses data from the PepLab platform, comprising 2748 experimentally confirmed bioactive peptides distributed across 15 functional classes, including ACE inhibitors, antimicrobial, anticancer, antioxidant, toxins, and others. For each peptide, the amino acid sequence and key physicochemical descriptors are provided, calculated via the integrated DMPep module, such as GRAVY index, aliphatic index, isoelectric point, molecular weight, Boman index, and sequence length. The dataset exhibits class imbalance, with class sizes ranging from 14 to 524 peptides. An innovative methodology is proposed, combining descriptive statistical analysis, structural modeling via DEMATEL, and structural equation modeling with neural networks (SEM-NN), where SEM-NN is used to capture complex nonlinear causal relationships between descriptors and functional classes. The results of these dependencies are integrated into a multi-class machine learning model to improve interpretability and predictive performance. Targeted data augmentation was applied to mitigate class imbalance. The developed classifier achieved predictive accuracy of up to 66%, a relatively high value given the complexity of the problem and the limited dataset size. These results confirm that integrating structured dependency modeling with artificial intelligence is an effective approach for functional peptide classification and supports the rational design of novel bioactive molecules. Full article

In off-road scenes, segmentation targets exhibit significant scale progression due to perspective depth effects from oblique viewing angles, meaning that the size of the same target undergoes continuous, boundary-less progressive changes along a specific direction. This asymmetric variation disrupts the geometric symmetry of targets, causing traditional segmentation networks to face three key challenges: (1) inefficientcapture of continuous-scale features, where pyramid structures and multi-scale kernels struggle to balance computational efficiency with sufficient coverage of progressive scales; (2) degraded intra-class feature consistency, where local scale differences within targets induce semantic ambiguity; and (3) loss of high-frequency boundary information, where feature sampling operations exacerbate the blurring of progressive boundaries. To address these issues, this paper proposes the ProCo-NET framework for systematic optimization. Firstly, a Progressive Strip Convolution Group (PSCG) is designed to construct multi-level receptive field expansion through orthogonally oriented strip convolution cascading (employing symmetric processing in horizontal/vertical directions) integrated with self-attention mechanisms, enhancing perception capability for asymmetric continuous-scale variations. Secondly, an Offset-Frequency Cooperative Module (OFCM) is developed wherein a learnable offset generator dynamically adjusts sampling point distributions to enhance intra-class consistency, while a dual-channel frequency domain filter performs adaptive high-pass filtering to sharpen target boundaries. These components synergistically solve feature consistency degradation and boundary ambiguity under asymmetric changes. Experiments show that this framework significantly improves the segmentation accuracy and boundary clarity of multi-scale targets in off-road scene segmentation tasks: it achieves 71.22% MIoU on the standard RUGD dataset (0.84% higher than the existing optimal method) and 83.05% MIoU on the Freiburg_Forest dataset. Among them, the segmentation accuracy of key obstacle categories is significantly improved to 52.04% (2.7% higher than the sub-optimal model). This framework effectively compensates for the impact of asymmetric deformation through a symmetric computing mechanism. Full article

The vibration signal of rotating machinery is usually nonlinear and non-stationary, and the feature set has information redundancy. Therefore, a high-dimensional feature reduction method based on multi-manifold learning is proposed for rotating machinery fault diagnosis. Firstly, considering the non-uniformity of multi-fault feature distribution and the sensitivity of domain selection in traditional manifold learning methods, the neighborhood size of each data point is selected adaptively by using the relationship between neighborhood size and sample density. Then, the between-manifold graph and within-manifold graph are constructed adaptively by the class information, and the divergence matrix and edge distance corresponding to the manifold graph are calculated. Feature fusion reduction is achieved by maximizing edge distance and minimizing within-class differences. Finally, the multi-manifold theoretical dataset and several rotating machinery fault datasets are selected for testing. The results show that the proposed algorithm has higher fault identification accuracy than traditional manifold learning methods. Full article

Forests, both natural and managed, provide a critical habitat for a significant part of global biodiversity. Among many different groups of forest biota, tree species occupy a special position as they create conditions upon which the existence of virtually all other forest organisms depends, either directly or indirectly. To permanently play this role, particular tree species must be demographically stable; i.e., their populations should be distinguished by the balanced, size-dependent proportions of individuals representing different developmental stages (from seedlings and saplings to mature and old trees). In this study, we examined the extent to which this condition is met in the managed part of Białowieża Forest in northeastern Poland, an important biodiversity hotspot in Central Europe. Comparison of species-specific equilibrium vs. actual size distributions revealed that almost half of all trees growing in Białowieża Forest represented “inappropriate” (i.e., occurring in excess compared to the balanced models) species and/or diameter ranges. The amount of deficits was also large (around 30% of the current tree number), concerning primarily the smallest trees. Considering this, we recommend targeted, active management strategies to restore the demographic balance of key tree species and, thus, to enhance the conservation of local biodiversity. We also indicate that the key elements of such strategies should be the gradual removal of trees from surplus diameter ranges and assisted regeneration of species with the greatest deficiencies in small diameter classes. Full article