Dataset Research Articles

Non-similar solution of Casson fluid flow over a curved stretching surface with viscous dissipation; Artificial neural network analysis

PurposeThe main focus is to provide a non-similar solution for the magnetohydrodynamic (MHD) flow of Casson fluid over a curved stretching surface through the novel technique of the artificial intelligence (AI)-based Lavenberg–Marquardt scheme of an artificial neural network (ANN). The effects of joule heating, viscous dissipation and non-linear thermal radiation are discussed in relation to the thermal behavior of Casson fluid.Design/methodology/approachThe non-linear coupled boundary layer equations are transformed into a non-linear dimensionless Partial Differential Equation (PDE) by using a non-similar transformation. The local non-similar technique is utilized to truncate the non-similar dimensionless system up to 2nd order, which is treated as coupled ordinary differential equations (ODEs). The coupled system of ODEs is solved numerically via bvp4c. The data sets are constructed numerically and then implemented by the ANN.FindingsThe results indicate that the non-linear radiation parameter increases the fluid temperature. The Casson parameter reduces the fluid velocity as well as the temperature. The mean squared error (MSE), regression plot, error histogram, error analysis of skin friction, and local Nusselt number are presented. Furthermore, the regression values of skin friction and local Nusselt number are obtained as 0.99993 and 0.99997, respectively. The ANN predicted values of skin friction and the local Nusselt number show stability and convergence with high accuracy.Originality/valueAI-based ANNs have not been applied to non-similar solutions of curved stretching surfaces with Casson fluid model, with viscous dissipation. Moreover, the authors of this study employed Levenberg–Marquardt supervised learning to investigate the non-similar solution of the MHD Casson fluid model over a curved stretching surface with non-linear thermal radiation and joule heating. The governing boundary layer equations are transformed into a non-linear, dimensionless PDE by using a non-similar transformation. The local non-similar technique is utilized to truncate the non-similar dimensionless system up to 2nd order, which is treated as coupled ODEs. The coupled system of ODEs is solved numerically via bvp4c. The data sets are constructed numerically and then implemented by the ANN.

Navigating GST revenue efficiency challenges: a solution to dilemma of policy makers for enhancing revenue efficiency

Purpose This study aims to investigate the determinants of GSTefficiency of the Indian states to assist the policymakers, government and GST council to devise their policies and strategies to boost the GSTefficiency of the Indian states. Design/methodology/approach The analysis has used the panel data set of 27 Indian states and 3 UTs with a time span of 2017–18 to 2022–23. The study has used the Generalized Method of Moment regression for exploring the determinants of GSTefficiency of the state governments in India. Findings The findings depict that sectoral composition, inflation rate, financial development, state’s self-reliance, per capita income and gross fiscal deficit have a significant effect on GSTefficiency of the state governments. The findings support the Tanzi effect 1977 and claim that the rise in the inflation level erodes GSTefficiency of the state governments. The rise in the self-reliance of the state government will make the Indian states self-dependent and will reduce their reliance on central transfers. Practical implications The government should make efforts to make the Indian states self-reliant by increasing the share of OTR (Own Tax Revenue) instead of increasing their revenue efficiency in short-run through devolution and central transfers. Moreover, the Indian government should devise their macro-economic policies to curb the inflation level and gross fiscal deficit of the state governments in the country. Originality/value To the best of the authors’ knowledge, this may be the first study to explore the determinants of GSTefficiency of the state governments in India.

Observation ofγγ→ττin proton-proton collisions and limits on the anomalous electromagnetic moments of the τ lepton.

The production of a pair of τ leptons via photon-photon fusion,γγ→ττ, is observed for the first time in proton-proton collisions, with a significance of 5.3 standard deviations. This observation is based on a data set recorded with the CMS detector at the LHC at a center-of-mass energy of 13 TeV and corresponding to an integrated luminosity of 138 fb-1. Events with a pair of τ leptons produced via photon-photon fusion are selected by requiring them to be back-to-back in the azimuthal direction and to have a minimum number of charged hadrons associated with their production vertex. The τ leptons are reconstructed in their leptonic and hadronic decay modes. The measured fiducial cross section ofγγ→ττisσobsfid=12.4-3.1+3.8fb. Constraints are set on the contributions to the anomalous magnetic moment (aτ) and electric dipole moments (dτ) of the τ lepton originating from potential effects of new physics on theγττvertex:aτ=0.0009-0.0031+0.0032and|dτ|<2.9×10-17ecm(95% confidence level), consistent with the standard model.

Exploring the human genomic landscape: patterns of common homozygosity regions in a large middle eastern cohort.

Regions of Homozygosity (ROH) typically reflect normal demographic history of a human population, but may also relate to cryptic consanguinity, and, additionally, have been associated with specific medical conditions. The objective of this study was to investigate the location, size, and prevalence of common ROH segments in a Middle Eastern cohort. This retrospective study included 13 483 samples collected from all Chromosomal Microarray analyses (CMA) performed using Single Nucleotide Polymorphism (SNP) arrays at the genetic clinical laboratory of Rabin Medical Center between 2017-2023 (primary data set). An additional replication cohort including 100 842 samples from another SNP array platform, obtained from Maccabi Health Organization, was analyzed. Common ROH locations were defined as those ROH locations involving 1% or more of the samples. A total of 66 710 ROH segments, involving 13 035 samples (96.7%) were identified in the primary data set. Of the 4069 cytogenetic ROH locations, 68 were identified as common. The prevalence of non-common ROH was relatively high in affected individuals, and for acrocentric chromosomes, chromosomes associated with common trisomies, and non-imprinted chromosomes. In addition, differences in common ROH locations were observed between the primary and the replication cohorts. Our findings highlight the need for population-specific guidelines in determining ROH reporting cutoffs, considering factors such as population-specific prevalence and testing platform differences. Future research with larger, varied cohorts is essential to advance understanding of ROH's associations with medical conditions and to improve clinical practices accordingly.

The interpretable descriptors for fatigue performance of wrought aluminum alloys

Aluminum alloys are extensively utilized in the transportation and aerospace industries due to their excellent processability and corrosion resistance. A key aspect affecting the application of aluminum alloys is fatigue failure, with fatigue strength constituting an essential indicator for assessing the failure mode. In this study, 859 data sets were collected to evaluate the effects of aluminum alloy composition and mechanical properties on fatigue strength. To comprehensively evaluate the characteristics that exert a great influence on the fatigue strength of aluminum alloys, the atomic physical quantities of aluminum alloys were added to the original data set. To improve the computational efficiency and physical interpretability of the model, Spearman correlation analysis, optimal subset method, and other methods are used to perform feature preprocessing on the atomic feature quantities and mechanical properties in the data set. Finally, a mathematical relationship between material descriptors and fatigue strength was established based on the alloy composition and the three most relevant features in the feature screening, which helps to better understand and predict the fatigue behavior of wrought aluminum alloys.

Deep neural networks and conventional machine learning classifiers to analyze thoracic survival data

Lung cancer is a prevalent global health concern and most prevalent malignancy in Indonesian hospitals. Following thoracic surgery, patients were categorized into two classes: individuals who experienced mortality within a year and those who achieved survival. Despite being about socks, the dataset for the deceased category consisted of 70 data samples, while the dataset for the final group comprised 400 samples. Data calculation involves the utilization of both deep neural networks and standard machine learning algorithms. The study use the Python programming language to evaluate the algorithms, and it measures their performance using metrics such as accuracy, F1-Score, precision, recall, receiver operating characteristic (ROC), and area under curve (AUC). The test results indicate that the deep neural network method achieves an accuracy of 95,56%, an F1 score of 79,24%, a precision of 91,96%, a recall of 85,52%, and an AUC of 85,52%. This study suggests that utilizing deep neural network data mining techniques, specifically with a cross-validation fold of 10, variations of six hidden layer encoder-decoder, relu, sigmoid activation function, optimizer Adam, and learning rate of 0,01, dropout rate of 0,2. Employing the Synthetic Minority Over-sampling Technique data preprocessing method, can effectively analyze thoracic patient survival data sets.

Bayesian ex Post Evaluation of Recursive Multi-Step-Ahead Density Prediction

This research is focused on a formal Bayesian method of recursive multi-step-ahead density prediction and its ex post evaluation. Our approach remains within the framework of the standard (classical or orthodox) Bayesian paradigm based on the Bayes factor and on the use of the likelihood-based update. We propose a new decomposition of the predictive Bayes factor into the product of partial Bayes factors, for both a finite number of consecutive k-step-ahead forecasts (where k>1) and the recursive updates of the posterior odds ratios based on updated data sets. The first factor in the decomposition is related to the relative k-step-ahead forecasting ability of models, while the second one measures the updating effect. To illustrate the usefulness of the proposed measures, we apply the new decomposed predictive Bayes factors to compare the forecasting ability of models when the true data generating process (DGP) is known, using simulated data sets. Taking into account the effect of updating, the posterior odds ratios leads to the conclusion that the best model coincides with the true DGP. However, the highest k-step-ahead forecasting ability (considered alone) can be achieved by some other, less adequate models. Next, we investigate the predictive ability of different Vector Error Correction (VEC) models with conditional heteroscedasticity, combining three macroeconomic variables: unemployment, inflation and interest rates, separately for the US and Polish economies. The results show that the inference about the models’ predictive performance depends on the forecast horizon as well as on taking into account the updating effect.

Early macular edema recognition after cataract surgery based on U-Net assisted optical coherence tomography

ObjectiveThe optical coherence tomography (OCT) model of macular edema in postoperative cataract patients is constructed by using convolutional neural network (CNN) U-Net, and its value is evaluated by relevant indicators. MethodsFrom January 2020 to December 2023, 200 binocular OCT images of 140 patients with macular edema after cataract surgery and 280 images of normal subjects in our hospital were included as data sets. They were randomly divided into a training set (70%) and valiadation set (30%). Res-SE U-Net was used to construct an automatic recognition model of macular edema in OCT images of patients after cataract surgery. The model was evaluated by the change of loss function and accuracy, and by drawing the receiver operating characteristic curve (ROC). ResultsThe segmentation efficacy between the U-Net model and its enhanced counterpart, Res-SE U-Net, was assessed in this study. Notably, the Dice coefficient for macular edema segmentation exhibited substantial improvement with the Res-SE U-Net model, reaching 0.96 compared to 0.90 for the conventional U-Net model. Additionally, evaluation based on the intersection over union (IoU) coefficient revealed a superiority of the Res-SE U-Net model with a coefficient of 0.79, in contrast to 0.76 for the U-Net model. Moreover, the area under the curve (AUC) analysis further supported the enhanced performance of the Res-SE U-Net model. The AUC values were determined as 0.837 for the Res-SE U-Net model and 0.811for the U-Net model. ConclusionThe Res-SE U-Net model of convolutional neural network based on OCT images can effectively identify and segment macular edema in patients after phacoemulsification.

BERT-based models for classifying multi-dialect Arabic texts

&lt;p&gt;&lt;a name="_Hlk133997549"&gt;&lt;/a&gt;&lt;a name="_Hlk127607325"&gt;&lt;/a&gt;&lt;span lang="EN-US"&gt;The area of natural language processing (NLP) is presently a rapidly developing field characterized by innovation and research. Despite this progress, several &lt;/span&gt;&lt;a name="_Hlk148089340"&gt;&lt;/a&gt;&lt;span lang="EN-US"&gt;dialects of Arabic&lt;/span&gt;&lt;span&gt;&lt;span lang="EN-US"&gt; (DA) are classified as low-resource languages, making it challenging for NLP systems to process DA data. One approach to address this issue is to train NLP models on social media data sets containing DA texts. &lt;/span&gt;&lt;span lang="IN"&gt;Therefore, these open-access social media datasets, as outlined in our paper, can serve as a valuable resource for developers and researchers involved in the processing of DA.&lt;/span&gt;&lt;span lang="EN-US"&gt;To create our multilingual corpus, we gathered data from various datasets containing different versions of DA. These datasets will be used to classify texts in terms of sentiment classification, topic classification, and dialect identification. Our study contributes to the automated analysis of the classification of Arabic dialects. We aim to investigate and assess various machine learning and deep learning techniques, with a specific focus on utilizing the BERT model. &lt;/span&gt;&lt;span lang="EN-US"&gt;The results of our experiments on our datasets show that DarijaBERT and DziriBERT trained on a similar DA outperform traditional machine learning methods and previous more general pre-trained models that were trained on multiple dialects or languages.&lt;/span&gt;&lt;/span&gt;&lt;/p&gt;

Semantic Segmentation of Point Cloud Scene via Multi-Scale Feature Aggregation and Adaptive Fusion

Point cloud semantic segmentation is a key step in 3D scene understanding and analysis. In recent years, deep learning–based point cloud semantic segmentation methods have received extensive attention from researchers. Multi-scale neighborhood feature learning methods are suitable for inhomogeneous density point clouds, but different scale branching feature learning increases the computational complexity and makes it difficult to accurately fuse different scale features to express local information. In this study, a point cloud semantic segmentation network based on RandLA-Net with multi-scale local feature aggregation and adaptive fusion is proposed. The designed structure can reduce computational complexity and accurately express local features. The mean intersection-over-union is improved by 1.1% on the SemanticKITTI data set with an inference speed of nine frames per second, while the mean intersection-over-union is improved by 0.9% on the S3DIS data set, compared with RandLA-Net. We also conduct ablation studies to validate the effectiveness of the proposed key structure.

Technology Focus: Reservoir Surveillance (September 2024)

In today’s oil and gas operations, surveillance technologies have undergone a revolution, reshaping how facilities, wells, and reservoirs are monitored. These advancements not only have increased the scale at which these technologies are deployed but also have led to an unprecedented influx of data. The sheer volume of data, however, poses a significant challenge to traditional analytical methods, overwhelming their capacity to derive actionable insights effectively. To address this challenge, the industry is rapidly advancing toward automated solutions powered by artificial-intelligence (AI) -driven analytics. These systems automate data ingestion and use machine-learning algorithms to sift through massive data sets, identifying anomalies and prioritizing actionable insights. By automating routine surveillance tasks, engineers can focus on critical actions that deliver substantial operational benefits. For example, in paper IPTC 23912, operators successfully optimized production operations by harnessing real-time field data through smart systems, effectively managing operations with complex near-critical fluids. Similarly, in paper SPE 218470, researchers proposed a novel workflow integrating virtual flowmetering and permanent downhole gauge data for pattern recognition to enhance real-time monitoring and decision-making in petroleum and geothermal industries. Nevertheless, ensuring effective surveillance of operational assets requires a strategic approach. A valuable tool in crafting such strategies is the value of information (VOI) assessment. This method systematically evaluates how acquiring specific information can influence decision-making and operational outcomes. For instance, paper SPE 215318 highlights a field operator’s systematic approach to VOI assessment, aiming to optimize daily operations and guide future development activities. In essence, while surveillance technologies have inundated operators with unprecedented data flows, advancements in automation and AI-driven analytics offer the promise of unlocking this data’s true potential. By embracing these technologies, the oil and gas industry can navigate the complexities of the modern energy landscape with greater agility, precision, and cost-effectiveness. Recommended additional reading at OnePetro: www.onepetro.org. SPE 215119 Surveillance, Analysis, and Optimization During Active Drilling Campaign by Yanfen Zhang, Chevron, et al. OTC 35413 New Opportunities in Well and Reservoir Surveillance Using Multiple Downhole Pressure Gauges in Deepwater Injector Wells by Piyush Pankaj, ExxonMobil, et al. OTC 34863 Digital Twin for Oil-Rim Management Using Early Warning System and Exception-Based Surveillance, Offshore Malaysia by M. Mahamad Amir, Petronas, et al.

A Self-Supervised Network-Based Smoke Removal and Depth Estimation for Monocular Endoscopic Videos.

In minimally invasive surgery videos, label-free monocular laparoscopic depth estimation is challenging due to smoke. For this reason, we propose a self-supervised collaborative network-based depth estimation method with smoke-removal for monocular endoscopic video, which is decomposed into two steps of smoke-removal and depth estimation. In the first step, we develop a de-endoscopic smoke for cyclic GAN (DS-cGAN) to mitigate the smoke components at different concentrations. The designed generator network comprises sharpened guide encoding module (SGEM), residual dense bottleneck module (RDBM) and refined upsampling convolution module (RUCM), which restores more detailed organ edges and tissue structures. In the second step, high resolution residual U-Net (HRR-UNet) consisting of a DepthNet and two PoseNets is designed to improve the depth estimation accuracy, and adjacent frames are used for camera self-motion estimation. In particular, the proposed method requires neither manual labeling nor patient computed tomography scans during the training and inference phases. Experimental studies on the laparoscopic data set of the Hamlyn Centre show that our method can effectively achieve accurate depth information after net smoking in real surgical scenes while preserving the blood vessels, contours and textures of the surgical site. The experimental results demonstrate that the proposed method outperforms existing state-of-the-art methods in effectiveness and achieves a frame rate of 94.45fps in real time, making it a promising clinical application.

Machine Learning Approach for Predicting Bead Geometry of Stainless Steel in Wire arc Additive Manufacturing

Wire arc additive manufacturing (WAAM) employs an electric arc to melt wire feedstock, making it a method within additive manufacturing (AM). It deposits material layer by layer to build up a part. The present study investigated the application of machine learning classification-based models for estimating bead width and bead height of stainless-steel parts fabricated using WAAM. The input parameters (voltage, current, wire feed rate, and travel speed) were considered as input to algorithms. Training and testing were performed for 98 experimental data sets from peer-reviewed literature. The machine learning classification models, K-nearest neighbors, decision tree with gini index as criteria, and random forest were evaluated. The ML model performance was evaluated utilizing statistical metrics, including accuracy, F1 score, precision, and recall. The decision tree classifier exhibited the highest accuracy of 87.8% for bead width and 84.7% for bead height. The findings offer valuable insights into leveraging ML techniques to enhance the performance and accuracy of predictive models within WAAM-based AM.

Surface Quasi Geostrophic Reconstruction of Vertical Velocities and Vertical Heat Fluxes in the Southern Ocean: Perspectives for SWOT

AbstractMesoscale currents account for 80% of the ocean's kinetic energy, whereas submesoscale currents capture 50% of the vertical velocity variance. SWOT's first sea surface height (SSH) observations have a spatial resolution an order of magnitude greater than traditional nadir‐looking altimeters and capture mesoscale and submesoscale features. This enables the derivation of submesoscale vertical velocities, crucial for the vertical transport of heat, carbon and nutrients between the ocean interior and the surface. This work focuses on a mesoscale energetic region south of Tasmania using a coupled ocean‐atmosphere simulation at km‐scale resolution and preliminary SWOT SSH observations. Vertical velocities (w), temperature anomalies and vertical heat fluxes (VHF) from the surface down to 1,000 m are reconstructed using effective surface Quasi‐Geostrophic (sQG) theory. An independent method for reconstructing temperature anomalies, mimicking an operational gridded product, is also developed. Results show that sQG reconstructs 90% of the modeled w and VHF rms at scales down to 30 km just below the mixed layer and 50%–70% of the rms for scales larger than 70 km at greater depth, with a spatial correlation of 0.6. The reconstruction is spectrally coherent for scales larger than 30–40 km at the surface, slightly degrading ( 0.55) at depth. Two temperature anomaly data sets yield similar results, indicating the dominance of w on VHF. The RMS of sQG and VHF derived from SWOT are twice as large as those derived from conventional altimetry, highlighting the potential of SWOT for reconstructing energetic meso and submesoscale dynamics in the ocean interior.

Risk prediction of cardiovascular disease in the Asia-Pacific region: the SCORE2 Asia-Pacific model.

To improve upon the estimation of 10-year cardiovascular disease (CVD) event risk for individuals without prior CVD or diabetes mellitus in the Asia-Pacific region by systematic recalibration of the SCORE2 risk algorithm. The sex-specific and competing risk-adjusted SCORE2 algorithms were systematically recalibrated to reflect CVD incidence observed in four Asia-Pacific risk regions, defined according to country-level World Health Organization age- and sex-standardized CVD mortality rates. Using the same approach as applied for the original SCORE2 models, recalibration to each risk region was completed using expected CVD incidence and risk factor distributions from each region. Risk region-specific CVD incidence was estimated using CVD mortality and incidence data on 8,405,574 individuals (556,421 CVD events). For external validation, data from 9,560,266 individuals without previous CVD or diabetes were analysed in 13 prospective studies from 12 countries (350,550 incident CVD events). The pooled C-index of the SCORE2 Asia-Pacific algorithms in the external validation data sets was 0.710 (95% confidence interval [CI] 0.677-0.745). Cohort-specific C-indices ranged from 0.605 (95% CI 0.597-0.613) to 0.840 (95% CI 0.771-0.909). Estimated CVD risk varied several-fold across Asia-Pacific risk regions. For example, the estimated 10-year CVD risk for a 50-year-old non-smoker, with a systolic blood pressure of 140 mmHg, total cholesterol of 5.5 mmol/L, and high-density lipoprotein-cholesterol of 1.3 mmol/L, ranged from 7% for men in low-risk countries to 14% for men in very-high-risk countries, and from 3% for women in low-risk countries to 13% for women in very-high-risk countries. The SCORE2 Asia-Pacific algorithms have been calibrated to estimate 10-year risk of CVD for apparently healthy people in Asia and Oceania, thereby enhancing the identification of individuals at higher risk of developing CVD across the Asia-Pacific region.

Stock market index prediction based on market trend using LSTM

The stock market data analysis has received interest as a result of technological advancements and the investigation of new machine learning models, since these models provide a platform for traders and business people to choose gaining stocks. The business price prediction is a challenging and extremely complex process due to the impact of several factors on company prices. The numerous patterns that the stock market goes, they have been the focus of extensive research and analysis by numerous experts. There are several large data sets accessible, an artificial intelligence and machine learning techniques are developing quickly, and because of the machine’s improved computational power, complex stock price prediction algorithms can be developed. This paper presents stock market index prediction based on market trend using long short-term memory (LSTM). Using built-in application programmable interface (API), Yahoo Finance offers a simple method to programmatically retrieve any historical stock prices of an organization using the ticker name. The standard and poor’s 500 index (S&amp;amp;P 500 index) include the firms that have been taken into consideration here. Utilizing the selected input variable, single-layer and multi-layer LSTM models are implemented, and the measurement parameters of mean absolute error (MAE), root mean square error (RMSE), and correlation coefficient (R) are used to compare each performance. Nearly all of the real closing price’s curve and the prediction curve’s closing price for test data overlap. A potential stock investor may benefit significantly from such a prediction by using it to make well-informed choices that would increase his earnings.

Performance Analysis of Cooperative Spectrum Sensing using Empirical Mode Decomposition and Artificial Neural Network in Wireless Regional Area Network

Background: Radio spectrum is natural and the most precious means in wireless communication systems. Optimal spectrum utilization is a key concern for today's cutting-edge wireless communication networks. The impending problem of the lack of available spectrum has prompted the development of a new idea called “Cognitive Radio” (CR). Cooperative spectrum sensing (CSS) is utilized to improve the detection performance of the system. Several fusion algorithms of decision-making are proposed for sensing the licensed user, but they do not work well under low signal-to-noise ratio (SNR). Objectives: To address the issue of poor detection performance under low SNR, Empirical mode decomposition (EMD) and artificial neural network (ANN) based CSS under Rayleigh multipath fading channel in IEEE 802.22 wireless regional area network (WRAN) is proposed in this paper. Method: In this work, we propose the use of ANN as a fusion center. First, the received signal's energy is calculated using EMD. The computed energy, SNR, and false alarm probability are combined to form a data set of 2048 samples. They are utilized to train Levenberg- Marquardt back propagation training algorithm-based feed-forward neural network (FFNN). Using this trained network, CSS in WRAN is simulated under Rayleigh multipath fading. Results: Simulation results show that the proposed CSS method based on EMD-ANN outperforms the standard fast Fourier transform (FFT) and EMD detection-based cooperative spectrum sensing with a hard "OR" fusion at low SNR. With Pf =0.01, 100% detection accuracy with proposed techniques is obtained at SNR= -22dB. Conclusion: The findings show that the suggested approach outperforms EMD and FFT based energy detection scheme-based traditional CSS in low SNR environments.

Human Orthohantavirus disease prevalence and genotype distribution in the U.S., 2008–2020: a retrospective observational study

In the United States (U.S.), hantavirus pulmonary syndrome (HPS) and non-HPS hantavirus infection are nationally notifiable diseases. Criteria for identifying human cases are based on clinical symptoms (HPS or non-HPS) and acute diagnostic results (IgM+, rising IgG+ titers, RT-PCR+, or immunohistochemistry (IHC)+). Here we provide an overview of diagnostic testing and summarize human Hantavirus disease occurrence and genotype distribution in the U.S. from 2008 to2020. Epidemiological data from the national hantavirus registry was merged with laboratory diagnostic testing results performed at the CDC. Residual hantavirus-positive specimens were sequenced, and the available epidemiological and genetic data sets were linked to conduct a genomic epidemiological study of hantavirus disease in the U.S. From 1993 to 2020, 833 human hantavirus cases have been identified, and from 2008 to 2020, 335 human cases have occurred. Among New World (NW) hantavirus cases detected at the CDC diagnostic laboratory (representing 29.2% of total cases), most (85.0%) were detected during acute disease, however, some convalescent cases were detected in states not traditionally associated with hantavirus infections (Connecticut, Missouri, New Jersey, Pennsylvania, Tennessee, and Vermont). From 1993 to 2020, 94.9% (745/785) of U.S. hantaviruses cases were detected west of the Mississippi with 45.7% (359/785) in the Four Corners region of the U.S. From 2008 to 2020, 67.7% of NW hantavirus cases were detected between the months of March and August. Sequencing of RT-PCR-positive cases demonstrates a geographic separation of Orthohantavirus sinnombreense species [Sin Nombre virus (SNV), New York virus, and Monongahela virus]; however, there is a large gap in viral sequence data from the Northwestern and Central U.S. Finally, these data indicate that commercial IgM assays are not concordant with CDC-developed assays, and that "concordant positive" (i.e., commercial IgM+ and CDC IgM+ results) specimens exhibit clinical characteristics of hantavirus disease. Hantaviral disease is broadly distributed in the contiguous U.S, viral variants are localised to specific geographic regions, and hantaviral disease infrequently detected in most Southeastern states. Discordant results between two diagnostic detection methods highlight the need for an improved standardised testing plan in the U.S. Hantavirus surveillance and detection will continue to improve with clearly defined, systematic reporting methods, as well as explicit guidelines for clinical characterization and diagnostic criteria. This work was funded by core funds provided to the Viral Special Pathogens Branch at CDC.

Bayesian Optimal Two-Sample Tests for High-Dimensional Gaussian Populations

We propose minimax optimal Bayesian two-sample tests for testing equality of high-dimensional mean vectors and covariance matrices between two populations. In many applications including genomics and medical imaging, it is natural to assume that only a few entries of two mean vectors or covariance matrices are different. Many existing tests that rely on aggregating the difference between empirical means or covariance matrices are not optimal or yield low power under such setups. Motivated by this, we develop Bayesian two-sample tests employing a divide-and-conquer idea, which is powerful especially when the differences between two populations are rare but large. The proposed two-sample tests manifest closed forms of Bayes factors and allow scalable computations even in high-dimensions. We prove that the proposed tests are consistent under relatively mild conditions compared to existing tests in the literature. Furthermore, the testable regions from the proposed tests turn out to be minimax optimal in terms of rates. Simulation studies show clear advantages of the proposed tests over other state-of-the-art methods in various scenarios. Our tests are also applied to the analysis of the gene expression data of two cancer data sets.

Gaussian Process-Gated Hierarchical Mixtures of Experts.

In this article, we propose novel Gaussian process-gated hierarchical mixtures of experts (GPHMEs). Unlike other mixtures of experts with gating models linear in the input, our model employs gating functions built with Gaussian processes (GPs). These processes are based on random features that are non-linear functions of the inputs. Furthermore, the experts in our model are also constructed with GPs. The optimization of the GPHMEs is performed by variational inference. The proposed GPHMEs have several advantages. They outperform tree-based HME benchmarks that partition the data in the input space, and they achieve good performance with reduced complexity. Another advantage is the interpretability they provide for deep GPs, and more generally, for deep Bayesian neural networks. Our GPHMEs demonstrate excellent performance for large-scale data sets, even with quite modest sizes.