Synthetic Data Research Articles

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.

Pop-cosmos: A Comprehensive Picture of the Galaxy Population from COSMOS Data

We present pop-cosmos: a comprehensive model characterizing the galaxy population, calibrated to 140,938 (r < 25 selected) galaxies from the Cosmic Evolution Survey (COSMOS) with photometry in 26 bands from the ultraviolet to the infrared. We construct a detailed forward model for the COSMOS data, comprising: a population model describing the joint distribution of galaxy characteristics and its evolution (parameterized by a flexible score-based diffusion model); a state-of-the-art stellar population synthesis model connecting galaxies’ intrinsic properties to their photometry; and a data model for the observation, calibration, and selection processes. By minimizing the optimal transport distance between synthetic and real data, we are able to jointly fit the population and data models, leading to robustly calibrated population-level inferences that account for parameter degeneracies, photometric noise and calibration, and selection. We present a number of key predictions from our model of interest for cosmology and galaxy evolution, including the mass function and redshift distribution; the mass–metallicity-redshift and fundamental metallicity relations; the star-forming sequence; the relation between dust attenuation and stellar mass, star formation rate, and attenuation-law index; and the relation between gas-ionization and star formation. Our model encodes a comprehensive picture of galaxy evolution that faithfully predicts galaxy colors across a broad redshift (z < 4) and wavelength range.

Coagulo-Net: Enhancing the mathematical modeling of blood coagulation using physics-informed neural networks

Blood coagulation, which involves a group of complex biochemical reactions, is a crucial step in hemostasis to stop bleeding at the injury site of a blood vessel. Coagulation abnormalities, such as hypercoagulation and hypocoagulation, could either cause thrombosis or hemorrhage, resulting in severe clinical consequences. Mathematical models of blood coagulation have been widely used to improve the understanding of the pathophysiology of coagulation disorders, guide the design and testing of new anticoagulants or other therapeutic agents, and promote precision medicine. However, estimating the parameters in these coagulation models has been challenging as not all reaction rate constants and new parameters derived from model assumptions are measurable. Although various conventional methods have been employed for parameter estimation for coagulation models, the existing approaches have several shortcomings. Inspired by the physics-informed neural networks, we propose Coagulo-Net, which synergizes the strengths of deep neural networks with the mechanistic understanding of the blood coagulation processes to enhance the mathematical models of the blood coagulation cascade. We assess the performance of the Coagulo-Net using two existing coagulation models with different extents of complexity. Our simulation results illustrate that Coagulo-Net can efficiently infer the unknown model parameters and dynamics of species based on sparse measurement data and data contaminated with noise. In addition, we show that Coagulo-Net can process a mixture of synthetic and experimental data and refine the predictions of existing mathematical models of coagulation. These results demonstrate the promise of Coagulo-Nets in enhancing current coagulation models and aiding the creation of novel models for physiological and pathological research. These results showcase the potential of Coagulo-Net to advance computational modeling in the study of blood coagulation, improving both research methodologies and the development of new therapies for treating patients with coagulation disorders.

SEMG-Based hand gestures classification using a semi-supervised multi-layer neural networks with Autoencoder

This work presents a semi-supervised multilayer neural network (MLNN) with an Autoencoder to develop a classification model for recognizing hand gestures from electromyographic (EMG) signals. Using a Myo armband equipped with eight non-invasive surface-mounted biosensors, raw surface EMG (sEMG) sensor data were captured corresponding to five hand gestures: Fist, Open hand, Wave in, Wave out, and Double tap. The sensor collected data underwent preprocessing, feature extraction, label assignment, and dataset organization for classification tasks. The model implementation, validation, and testing demonstrated its efficacy after incorporating synthetic sEMG data generated by an Autoencoder. In comparison to the state-of-the-art techniques from the literature, the proposed model exhibited strong performance, achieving accuracy of 99.68%, 100%, and 99.26% during training, validation, and testing, respectively. Comparatively, the proposed MLNN with Autoencoder model outperformed a K-Nearest Neighbors model established for comparative evaluation.

Hierarchical homogeneity-based superpixel segmentation: application to hyperspectral image analysis

ABSTRACT Hyperspectral image (HI) analysis approaches have recently become increasingly complex and sophisticated. Recently, the combination of spectral-spatial information and superpixel techniques have addressed some hyperspectral data issues, such as the higher spatial variability of spectral signatures and dimensionality of the data. However, most existing superpixel approaches do not account for specific HI characteristics resulting from its high spectral dimension. In this work, we propose a multiscale superpixel method that is computationally efficient for processing hyperspectral data. The Simple Linear Iterative Clustering (SLIC) oversegmentation algorithm, on which the technique is based, has been extended hierarchically. Using a novel robust homogeneity testing, the proposed hierarchical approach leads to superpixels of variable sizes but with higher spectral homogeneity when compared to the classical SLIC segmentation. For validation, the proposed homogeneity-based hierarchical method was applied as a preprocessing step in the spectral unmixing and classification tasks carried out using, respectively, the Multiscale sparse Unmixing Algorithm (MUA) and the CNN-Enhanced Graph Convolutional Network (CEGCN) methods. Simulation results with both synthetic and real data show that the technique is competitive with state-of-the-art solutions.

Denoising graph neural network based on zero-shot learning for Gibbs phenomenon in high-order DG applications

With the availability of high-performance computing technology and the development of advanced numerical simulation methods, Computational Fluid Dynamics (CFD) is becoming more and more practical and efficient in engineering. As one of the high-precision representative algorithms, the high-order Discontinuous Galerkin Method (DGM) has not only attracted widespread attention from scholars in the CFD research community, but also received strong development. However, when DGM is extended to high-speed aerodynamic flow field calculations, non-physical numerical Gibbs oscillations near shock waves often significantly affect the numerical accuracy and even cause calculation failure. Data driven approaches based on machine learning techniques can be used to learn the characteristics of Gibbs noise, which motivates us to use it in high-speed DG applications. To achieve this goal, labeled data need to be generated in order to train the machine learning models. This paper proposes a new method for denoising modeling of Gibbs phenomenon using a machine learning technique, the zero-shot learning strategy, to eliminate acquiring large amounts of CFD data. The model adopts a graph convolutional network combined with graph attention mechanism to learn the denoising paradigm from synthetic Gibbs noise data and generalize to DGM numerical simulation data. Numerical simulation results show that the Gibbs denoising model proposed in this paper can suppress the numerical oscillation near shock waves in the high-order DGM. Our work automates the extension of DGM to high-speed aerodynamic flow field calculations with higher generalization and lower cost.

An algorithm for simplified recurrence analysis.

Recurrence analysis applications are hindered by several issues including the selection of critical parameters, noise sensitivity, computational complexity, or the analysis of non-stationary systems. Great progresses have been made by the community to address these issues individually, yet the diversity of resulting techniques with often additional parameters as well as a lack of consensus still impedes its use by nonspecialists. We present a procedure for simplified recurrence analysis based on compact recurrence plots with automatized parameter selection and enhanced noise robustness, and that are suited to the analysis of complex non-stationary systems. This approach aims at supporting the expansion of recurrence analysis for currently challenging or future applications such as for large systems, on-site studies, or using machine learning. The method is demonstrated on both synthetic and real data showing promising results.

Subspace projection attention network: Deep weak signal recovery from complicated downhole distributed acoustic sensing data

Distributed acoustic sensing (DAS) is a new, rapidly developing acquisition technology. It has many advantages, such as full well coverage, high sampling density, and strong tolerance compared with conventional geophones in harsh environments. However, DAS data contain various new types of noise with different characteristics, which brings great difficulties to weak signal recovery. It is also not conducive to subsequent inversion, imaging, and interpretation. Several traditional methods and their developments can only remove particular types of DAS noise with serious energy loss. At the same time, the denoising performance is strongly dependent on the threshold function or manual setting parameters. In recent years, deep learning has attracted much attention in DAS data processing. However, convolutional neural networks can only capture local information. The commonly used channel attention assigns weights to each channel to reflect the importance and neglects the features within the channels. The popular self-attention mechanism (SAM) can efficiently obtain the dependencies between different locations or different patches that are a long distance from the input. Thus, a novel subspace projection attention network (SPANet) is designed from a new projection perspective. In the novel network, some feature basis vectors are generated to guide the decoupled projection of the noise components from the signal components. Low-level surface features are projected under the guidance of high-level abstract features through SAM before feature fusion. The projection operation helps to find the optimal subspace that preserves the fine structure of the input as much as possible, better capturing the potential relevance in DAS data. Finally, experimental results of synthetic and field data illustrate that SPANet recovers the effective reflection signals well, whether in a shallow layer with strong noise or a deep layer with weak energy. Its superior performance compared with some traditional and network methods is fully verified and analyzed.

Petrophysical parameter estimation using a rock-physics model and collaborative sparse representation

The estimation of petrophysical parameters is key in the identification of underground reservoirs. Current petrophysical parameter estimation methods are typically constrained by the choice of particular rock-physics models, necessitating the use of distinct models for various reservoirs. Furthermore, the inherent pronounced nonlinearity of these models presents significant challenges to the solution process in reservoir parameter inversion. Although linearized petrophysical inversion methods can simplify the solving process, they can introduce errors due to linearization. To address these limitations, we develop a petrophysical parameter estimation method driven by a rock-physics model and collaborative sparse representation (CSR). In this approach, the rock-physics model governs the fundamental pattern of the petrophysical parameters, with the CSR introducing perturbations to minimize model errors. Our method maximizes the use of well-log data and the rock-physics model, thereby reducing errors associated with linearized rock-physics (LRP) models and enhancing the method’s adaptability. We use the joint dictionary method to learn features and correlations among multiple parameters from the existing well-log data. This learned dictionary is then used as a CSR regularization constraint and applied to refine the LRP inversion objective function. Finally, the objective function is minimized using the block coordinate descent method to predict the petrophysical parameters. The effectiveness of this method in enhancing the adaptability and accuracy of petrophysical parameter estimation is confirmed through synthetic and field data tests.

Deep-learning optimization using the gradient of a custom objective function: A full-waveform inversion example study on the convolutional objective function

The integration of conventional high-performance full-waveform inversion (FWI) algorithms with deep-learning frameworks is an innovative and promising research direction with the potential to enhance and broaden the application prospects of this field significantly. Automatic differentiation with backpropagation techniques can derive the gradients of model parameters in an equivalent manner to that of adjoint methods; however, it requires a substantial amount of computer memory, particularly when considering the time-step layers. In addition, some excellent objective functions suitable for FWI are not available in deep-learning frameworks. In comparison, the adjoint method, based on effective boundary storage technology, offers greater practicality for calculating the gradients of various objective functions. Therefore, this paper develops a novel approach toward FWI that integrates deep-learning optimization with high-performance gradient computation. In particular, our method inputs model parameter gradients from a custom objective function into the deep-learning framework. Herein, we use the acoustic equation with variable density as an example to demonstrate how a convolutional objective function, along with its corresponding velocity and density gradients, can be used for optimized inversion, multiscale inversion, and deep network parameterization-based multiscale inversion within the deep-learning framework. This approach provides a paradigm for deep-learning-optimized FWI, which we apply to synthetic and field data scenarios.

Generative Variational-Contrastive Learning for Self-Supervised Point Cloud Representation.

Self-supervised representation learning for 3D point clouds has attracted increasing attention. However, existing methods in the field of 3D computer vision generally use fixed embeddings to represent the latent features, and impose hard constraints on the embeddings to make the latent feature values of the positive samples converge to consistency, which limits the ability of feature extractors to generalize over different data domains. To address this issue, we propose a Generative Variational-Contrastive Learning (GVC) model, where Gaussian distribution is used to construct a continuous, smoothed representation of the latent features. A distribution constraint and cross-supervision are constructed to improve the transfer ability of the feature extractor over synthetic and real-world data. Specifically, we design a variational contrastive module to constrain the feature distribution instead of feature values corresponding to each sample in the latent space. Moreover, a generative cross-supervision module is introduced to preserve the invariance features and promote the consistency of feature distribution among positive samples. Experimental results demonstrate that GVC achieves SOTA on different downstream tasks. In particular, with only pre-training on the synthetic dataset, GVC achieves a lead of 8.4% and 14.2% when transferring to the real-world dataset in the linear classification and few-shot classification.

KF-PEV: a causal Kalman filter-based particle event velocimetry

Event-based pixel sensors asynchronously report changes in log-intensity in microsecond-order resolution. Its exceptional speed, cost effectiveness, and sparse event stream make it an attractive imaging modality for particle tracking velocimetry. In this work, we propose a causal Kalman filter-based particle event velocimetry (KF-PEV). Using the Kalman filter model to track the events generated by the particles seeded in the flow medium, KF-PEV yields the linear least squares estimate of the particle track velocities corresponding to the flow vector field. KF-PEV processes events in a computationally efficient and streaming manner (i.e., causal and iteratively updating). Our simulation-based benchmarking study with synthetic particle event data confirms that the proposed KF-PEV outperforms the conventional frame-based particle image/tracking velocimetry as well as the state-of-the-art event-based particle velocimetry methods. In a real-world water tunnel event-based sensor data experiment performed on what we believe to be the widest field view ever reported, KF-PEV accurately predicted the expected flow field of the SD7003 wing, including details such as the lower velocity in the wake and the flow separation around the underside of an angled wing.

The use of digital thread for reconstruction of local fiber orientation in a compression molded pin bracket via deep learning

A deep convolutional neural network (DCNN) was used for microstructure reconstruction using artificial intelligence (MR-AI) by predicting local average fiber orientation distributions (FOD) in a 3D prepreg platelet molded composite (PPMC) pin bracket. To train the MR-AI model, surface strain fields from residual stresses simulated in PPMC plates were used as the input to the DCNN. A training dataset included PPMC plates with various degrees of global fiber alignment, based on the information obtained from high-fidelity flow simulation of a pin bracket. The MR-AI model was then deployed to analyze FOD in the 3D pin bracket by conducting thermo-elastic residual stress analysis. Initially, the MR-AI model was established entirely on the synthetic simulation data. Then, a μCT scan of a physically molded pin bracket was used to create a finite element model that provided data for additional validation of the DCNN model. For the μCT scan finite element pin bracket the MR-AI model predicted the distribution of fiber orientation tensor components with MAE of 0.10 indicating a global prediction error of 10 %. For the flow simulated pin bracket, the MR-AI model predicted the distribution of fiber orientation tensor components with a global prediction error of 11 %. The MR-AI model showed the ability to predict regions of varying alignment in the base and flange of the pin bracket. The proposed MR-AI methodology allows for rapid prediction of FOD in geometrically complex parts and offers a promising path to detecting unique fiber orientation states in molded components.

Online kernel-based clustering

A novel online joint kernel learning and clustering (OKC) framework is derived which is capable of determining time-varying clustering configurations without the need for training data. To facilitate clustering via sparse kernel factorization, a novel time-dependent metric is devised to quantify closeness of candidate kernel similarity matrices to a block diagonal structure. The task is later performed online as newly acquired data are processed making it appropriate for non-stationary settings. The necessary minimization to carry out kernel learning and clustering is performed via an effective interplay among block coordinate descent, difference of convex functions minimization and projected subgradient descent which results in an online iterative algorithm that updates online the kernel covariance matrix, as well as the associated clustering configuration. The resulting recursive kernel updates are proved to converge to a bounded limit. Detailed numerical examples utilizing both synthetic and real data show the superior clustering accuracy achieved by the novel approach over existing alternatives while requiring considerably less running time.

Advancing predictive accuracy of shallow landslide using strategic data augmentation

Rainfall-induced shallow landslides pose one of significant geological hazards, necessitating precise monitoring and prediction for effective disaster mitigation. Most studies on landslide prediction have focused on optimizing machine learning (ML) algorithms, very limited attention has been paid to enhancing data quality for improved predictive performance. This study employs strategic data augmentation (DA) techniques to enhance the accuracy of shallow landslide prediction. Using five DA methods including singular spectrum analysis (SSA), moving averages (MA), wavelet denoising (WD), variational mode decomposition (VMD), and linear interpolation (LI), we utilize strategies such as smoothing, denoising, trend decomposition, and synthetic data generation to improve the training dataset. Four machine learning algorithms, i.e. artificial neural network (ANN), recurrent neural network (RNN), one-dimensional convolutional neural network (CNN1D), and long short-term memory (LSTM), are used to forecast landslide displacement. The case study of a landslide in southwest China shows the effectiveness of our approach in predicting landslide displacements, despite the inherent limitations of the monitoring dataset. VMD proves the most effective for smoothing and denoising, improving R2, RMSE, and MAPE by 172.16%, 71.82%, and 98.9%, respectively. SSA addresses missing data, while LI is effective with limited data samples, improving metrics by 21.6%, 52.59%, and 47.87%, respectively. This study demonstrates the potential of DA techniques to mitigate the impact of data defects on landslide prediction accuracy, with implications for similar cases.

Explainable Artificial Intelligence for Early Prediction of Pressure Injury Risk.

Hospital-acquired pressure injuries (HAPIs) have a major impact on patient outcomes in intensive care units (ICUs). Effective prevention relies on early and accurate risk assessment. Traditional risk-assessment tools, such as the Braden Scale, often fail to capture ICU-specific factors, limiting their predictive accuracy. Although artificial intelligence models offer improved accuracy, their "black box" nature poses a barrier to clinical adoption. To develop an artificial intelligence-based HAPI risk-assessment model enhanced with an explainable artificial intelligence dashboard to improve interpretability at both the global and individual patient levels. An explainable artificial intelligence approach was used to analyze ICU patient data from the Medical Information Mart for Intensive Care. Predictor variables were restricted to the first 48 hours after ICU admission. Various machine-learning algorithms were evaluated, culminating in an ensemble "super learner" model. The model's performance was quantified using the area under the receiver operating characteristic curve through 5-fold cross-validation. An explainer dashboard was developed (using synthetic data for patient privacy), featuring interactive visualizations for in-depth model interpretation at the global and local levels. The final sample comprised 28 395 patients with a 4.9% incidence of HAPIs. The ensemble super learner model performed well (area under curve = 0.80). The explainer dashboard provided global and patient-level interactive visualizations of model predictions, showing each variable's influence on the risk-assessment outcome. The model and its dashboard provide clinicians with a transparent, interpretable artificial intelligence-based risk-assessment system for HAPIs that may enable more effective and timely preventive interventions.

Framework to improve software effort estimation accuracy using novel ensemble rule

This investigation focuses on refining software effort estimation (SEE) to enhance project outcomes amidst the rapid evolution of the software industry. Accurate estimation is a cornerstone of project success, crucial for avoiding budget overruns and minimizing the risk of project failures. The framework proposed in this article addresses three significant issues that are critical for accurate estimation: dealing with missing or inadequate data, selecting key features, and improving the software effort model. Our proposed framework incorporates three methods: the Novel Incomplete Value Imputation Model (NIVIM), a hybrid model using Correlation-based Feature Selection with a meta-heuristic algorithm (CFS-Meta), and the Heterogeneous Ensemble Model (HEM). The combined framework synergistically enhances the robustness and accuracy of SEE by effectively handling missing data, optimizing feature selection, and integrating diverse predictive models for superior performance across varying project scenarios. The framework significantly reduces imputation and feature selection overhead, while the ensemble approach optimizes model performance through dynamic weighting and meta-learning. This results in lower mean absolute error (MAE) and reduced computational complexity, making it more effective for diverse software datasets. NIVIM is engineered to address incomplete datasets prevalent in SEE. By integrating a synthetic data methodology through a Variational Auto-Encoder (VAE), the model incorporates both contextual relevance and intrinsic project features, significantly enhancing estimation precision. Comparative analyses reveal that NIVIM surpasses existing models such as VAE, GAIN, K-NN, and MICE, achieving statistically significant improvements across six benchmark datasets, with average RMSE improvements ranging from 11.05% to 17.72% and MAE improvements from 9.62% to 21.96%. Our proposed method, CFS-Meta, balances global optimization with local search techniques, substantially enhancing predictive capabilities. The proposed CFS-Meta model was compared to single and hybrid feature selection models to assess its efficiency, demonstrating up to a 25.61% reduction in MSE. Additionally, the proposed CFS-Meta achieves a 10% (MAE) improvement against the hybrid PSO-SA model, an 11.38% (MAE) improvement compared to the Hybrid ABC-SA model, and 12.42% and 12.703% (MAE) improvements compared to the hybrid Tabu-GA and hybrid ACO-COA models, respectively. Our third method proposes an ensemble effort estimation (EEE) model that amalgamates diverse standalone models through a Dynamic Weight Adjustment-stacked combination (DWSC) rule. Tested against international benchmarks and industry datasets, the HEM method has improved the standalone model by an average of 21.8% (Pred()) and the homogeneous ensemble model by 15% (Pred()). This comprehensive methodology underscores our model’s contributions to advancing software project management (SPM) through advanced predictive modeling, setting a new benchmark for software engineering effort estimation.

To democratize research with sensitive data, we should make synthetic data more accessible

To democratize research with sensitive data, we should make synthetic data more accessible

Machine Learning‐Assisted Microearthquake Location Workflow for Monitoring the Newberry Enhanced Geothermal System

AbstractEnhanced geothermal systems (EGS) offer a sustainable energy source but face challenges in accurately locating microearthquakes induced during reservoir stimulation. Locating these microearthquakes provides reliable feedback on the stimulation progress. Current deep learning methods for locating earthquakes require extensive data sets for training, which is problematic as detected microearthquakes are often limited. To address the scarcity of training data, we propose a practical workflow using probabilistic multilayer perceptron (PMLP) which predicts microearthquake locations from cross‐correlation time lags in waveforms. Utilizing a 3D velocity model of Newberry site derived from ambient noise interferometry, we generate numerous synthetic microearthquakes and 3D acoustic waveforms for PMLP training. Accurate synthetic tests prompt us to apply the trained network to the 2012 and 2014 stimulation field waveforms. To enhance the accuracy of source localization, we carefully handpick the P‐arrival times. Predictions on the 2012 stimulation data set show major microseismic activity at depths of 0.5–1.2 km, correlating with a known casing leakage scenario. In the 2014 data set, the majority of predictions concentrate at 2.0–2.9 km depths, consistent with results obtained from conventional physics‐based inversion, and align with the presence of natural fractures from 2.0 to 2.7 km. We validate our findings by comparing the synthetic and field picks, demonstrating a satisfactory match for the first arrivals. By combining the benefits of quick inference speeds and accurate location predictions, we demonstrate the feasibility of using realistic synthetic data set to locate microseismicity for EGS monitoring.

Reconstructing SARS-CoV-2 lineages from mixed wastewater sequencing data

Wastewater surveillance of SARS-CoV-2 has emerged as a critical tool for tracking the spread of COVID-19. In addition to estimating the relative case numbers using quantitative PCR, SARS-CoV-2 genomic RNA can be extracted from wastewater and sequenced. There are many existing techniques for using the sequenced RNA to determine the relative abundance of known lineages in a sample. However, it is very challenging to predict novel lineages from wastewater data due to its mixed composition and unreliable genomic coverage. In this work, we present a novel technique based on non-negative matrix factorization which is able to reconstruct lineage definitions by analyzing data from across different samples. We test the method both on synthetic and real wastewater sequencing data. We show that the technique is able to determine major lineages such as Omicron and Delta as well as sub-lineages such as BA.5.2.1. We provide a method for determining emerging lineages in wastewater without the need for genomic data from clinical samples. This could be used for routine monitoring of SARS-CoV-2 as well as other emerging viral pathogens in wastewater. Additionally, it may be used to determine more full-genome sequences for viruses with fewer available genomes.