Thresholds Of Network Research Articles

Determining structural sites of the IGPD protein family from patterns in property based correlation matrices

Abstract We use the statistical approach of random matrix and network theory to tackle the problem of identifying the important motifs responsible for the crucial functioning of the IGPD protein family. This addresses directly the question of patterns of interaction between amino acid residues (based on properties) in proteins that contribute to protein function. We use the mathematical tools of inverse participation ratio and Shannon entropy to determine the locations of the important groups of correlated amino acid positions which gives us the structural sites of the IGPD protein. These tools isolate the smallest eigenvalues/outliers corresponding to eigenmodes as the most localized which give the crucial sites for the family. We also create the threshold network of the IGPD protein and find that at certain threshold, similar sites emerge from the network analysis which in addition gives us the strongest connected sites. This strengthens our method of finding the structural and functional sites. As a bonus we find these important sites also match with experiments.

Tail risk network of Chinese green-related stocks market

This paper explores the tail risk network in the Chinese green-related stock market, by estimating the Copula-MIDAS-LASSO model. The model integrates characteristic factors and improves the model’s ability to capture the tail risk spillover. Additionally, we utilize the MTS network and the threshold network to picture the network in the market. The findings reveal that securities consistently hold central positions in the risk contagion network, with GFS, CUB, and PAI being three key sources of risk contagion. Finally, the subsample analysis demonstrates that the green finance policy and financial crises contribute to increased risk dependence within the market.

BRAIN TUMOR PERCEPTION USING DEEP LEARNING

The human brain serves as the central component of the human body, orchestrating its functionality. Brain tumors emerge from abnormal cell growth and division within the brain, potentially leading to the development of brain cancer. Computer vision plays a pivotal role in healthcare by alleviating the burden of making precise decisions. Among imaging modalities like magnetic resonance imaging (MRI), computed tomography (CT) scans, and X-rays, MRI scans stand out as one of the most prevalent and safest methods for obtaining detailed images, capable of detecting minute abnormalities. Our research endeavors to explore the diverse applications of brain MRI in detecting brain cancer. In this study, we employed the bilateral filter (BF) to eliminate noise from MR images before processing. Subsequently, the tumor region was delineated using binary thresholding and Convolutional Neural Network (CNN) segmentation techniques. The datasets were divided into training, testing, and validation subsets, enabling our machine to discern the presence of brain tumors in subjects. Various performance metrics including accuracy, sensitivity, and specificity were utilized to evaluate the outcomes. Dense layers within the CNN architecture efficiently extract features from brain MRI images, enhancing diagnostic accuracy. The experiment was conducted primarily on MRI data due to its ability to provide comprehensive insights into cellular

Pushing the Limits of Acoustic Spatial Perception via Incident Angle Encoding

With the growing popularity of smart speakers, numerous novel acoustic sensing applications have been proposed for low-frequency human speech and high-frequency inaudible sounds. Spatial information plays a crucial role in these acoustic applications, enabling various location-based services. However, typically commercial microphone arrays face limitations in spatial perception of inaudible sounds due to their sparse array geometries optimized for low-frequency speech. In this paper, we introduce MetaAng, a system designed to augment microphone arrays by enabling wideband spatial perception across both speech signals and inaudible sounds by leveraging the spatial encoding capabilities of acoustic metasurfaces. Our design is grounded in the fact that, while sensitive to high-frequency signals, acoustic metasurfaces are almost non-responsive to low-frequency speech due to significant wavelength discrepancy. This observation allows us to integrate acoustic metasurfaces with sparse array geometry, simultaneously enhancing the spatial perception of high-frequency and low-frequency acoustic signals. To achieve this, we first utilize acoustic metasurfaces and a configuration optimization algorithm to encode the unique features for each incident angle. Then, we propose an unrolling soft thresholding network that employs neural-enhanced priors and compressive sensing for high-accuracy, high-resolution multi-source angle estimation. We implement a prototype, and experimental results demonstrate that MetaAng maintains robustness across various scenarios, facilitating multiple applications, including localization and tracking.

Investigation of Indian stock markets using topological data analysis and geometry-inspired network measures

Financial markets are a quintessential example of complex systems wherein strategic connections arising through the trading (buying and selling) of different stocks gives rise to dynamic correlation patterns in stock prices. Network science based approaches have been employed for more than two decades to investigate the stock market as an inferential network, with nodes representing stocks and edges as pairwise cross-correlations between stock prices. However, a recent focus has been the characterization of the higher-order structure, beyond pairwise correlation, of such networks. To this end, geometry-inspired measures (such as discrete Ricci curvatures) and topological data analysis (TDA) based methods (such as persistent homology) have become attractive tools for characterizing the higher-order structure of networks representing the financial systems. In this study, our goal is to perform a comparative analysis of both these approaches, especially by assessing the fragility and systemic risk in the Indian stock markets, which is known for its high volatility and risk. To achieve this goal, we analyze the time series of daily log-returns of stocks comprising the National Stock Exchange (NSE) and the Bombay Stock Exchange (BSE). Specifically, our aim is to monitor the changes in standard network measures, edge-centric discrete Ricci curvatures, and persistent homology based topological measures computed from cross-correlation matrices of stocks. In this study, the edge-centric discrete Ricci curvatures have been employed for the first time in the analysis of the Indian stock markets. The Indian stock markets are known to be less diverse in comparison to the US market, and hence provides us an interesting example. Our results point that, among the persistent homology based topological measures, persistent entropy is simple and more robust than L1-norm and L2-norm of persistence landscape. In a broader comparison between network analysis and TDA, we highlight that the network analysis is sensitive to the way of constructing the networks (threshold or minimum spanning tree), as well as the threshold values used to construct the correlation-based threshold networks. On the other hand, the persistent homology is a more robust approach and is able to capture the higher-order interactions and eliminate noisy data in financial systems, since it does not take into account a single value of threshold but rather a range of values.

Adaptive thresholding and coordinate attention-based tree-inspired network for aero-engine bearing health monitoring under strong noise

In engineering, due to strong noise interference, it is a challenging issue for decision-making of the neural network to accurately detect healthy conditions of aero-engine bearings. As such, a hierarchical health monitoring model, termed adaptive thresholding and coordinate attention-based tree-inspired network (ATCATN) is developed for the health monitoring of aero-engine bearings under strong background noise. The ATCATN integrates the advantages of the adaptive thresholding and coordinate attention-based network (ATCAN) as well as a tree structure-based method, and it is not merely a straightforward integration of two models. First, the ATCAN is constructed as the backbone network, where the coordinate attention module is developed to process one-dimensional vibration signals; meanwhile, the designed deep residual adaptive thresholding module is introduced to extract significant features from vibration signals with strong noise. Second, a two-layer tree-inspired decision layer, consisting of seed and leaf nodes, is developed to reconstruct the output layer of the ATCAN. Finally, the trained ATCATN is more consistent with maintenance cognition involving multilevel decision information by the progressive determination of fault locations and fault sizes. The experimental results and comparative analysis with seven advanced methods on two aero-engine bearing datasets show that the developed model has the following advantages: (a) model robustness under strong noise; (b) more accurate diagnosis result and (c) hierarchical diagnosis with fault locations and fault sizes. Based on the experimental results, the ATCATN model demonstrates high accuracy in progressively identifying fault locations and sizes of aero-engine bearings even under strong noise interference, which is beneficial for formulating maintenance strategies and aligns with the basic understanding of maintenance.

Abstract 6878: Images to insights: Transforming TME characterization into actionable spatial metrics with AI-powered multiplex immunofluorescence

Abstract Background: Precise and actionable assessment of cellular mechanisms and interactions within the tumor microenvironment (TME) is critical for advancing clinical trials, preclinical research, and informed decision-making in immuno-oncology. Multiplex immunofluorescence (mIF) assays have emerged as pivotal tools that can enable comprehensive visualization of protein expression and cellular interactions within the TME. Integration of advanced digital analysis tools and mIF assays, can revolutionize our ability to interpret the TME and usher these assays into high-throughput clinical settings. Methods: In this study, we processed tumor samples utilizing Akoya's PhenoCode Signature Immune Profile Human Protein Panel, which includes immune profiling markers CD8, CD68, CD3, CD20 and tumor marker PanCK. Following the Akoya Opal staining protocol, formalin-fixed paraffin-embedded (FFPE) samples were stained and scanned using the PhenoImager HT scanner. The images were then analyzed using Reveal’s AI-mIF digital assay on the ImageDx platform to generate contextual and spatial biomarker data. A deep learning-based cell segmentation model based on MaskRCNN architecture, signal preprocessing for noise and autofluorescence, and dynamic intensity normalization were used. Phenotype classification was precisely executed using dynamic thresholding or neural network classifiers as needed. Results were validated by establishing concordance with pathologist readouts. Results: The application of our methodology yielded highly concordant results when compared with manual counts, demonstrating the efficacy of our approach. The R2 correlation coefficient across all markers passed the minimum criteria of R2>0.85 with regression slopes between 0.7-1.3, indicating a robust alignment with traditional pathologist assessments. We showcase stratifying samples Desert, Excluded, or Inflamed phenotypes and other quantitative and spatial results such as interactions of immune cells and tumor within 30-um neighborhoods, and their dispersion and arrangement in the TME. Conclusion: The integration of Reveal Biosciences' ImageDx platform and RevealAI-mIF provides an ecosystem to host, visualize, and analyze Akoya’s immune profiling images. This collaborative effort has enabled the transformation of complex multiplex imaging data into distinct spatial phenomic endpoints, thereby introducing the potential to significantly advance biomarker discovery and contribution to translational and clinical research in the field of immuno-oncology. Citation Format: Misagh Naderi, Lazar Krstic, Djordje Cikic, Kristen Ruf, Igor Mihajlovic, Sinisa Todorovic, Claire Weston. Images to insights: Transforming TME characterization into actionable spatial metrics with AI-powered multiplex immunofluorescence [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6878.

The Bearing Surface Defect Detection Method Combining Magnetic Particle Testing and Deep Learning

As a critical foundational component, bearings find widespread application in various mechanical equipment. In order to achieve automated defect detection in the bearing-manufacturing process, a defect detection algorithm combining magnetic particle inspection with deep learning is proposed. Dynamic thresholding and generative adversarial network (GAN) methods are employed to extract defect samples from bearing images and augment the dataset, thereby enhancing data diversity. To mitigate the impact of irrelevant displays in bearing images, a coordinated attention (CA) mechanism is introduced into the backbone network of the deep learning model to focus on key information. Additionally, an adaptive spatial feature fusion module (ASFF) is incorporated during the multiscale fusion stage to maintain consistency in features across different hierarchical levels. The weighted intersection over union (WIoU) bounding box loss function is utilized to replace the original generalized intersection over union (GIoU) in the network, directing the model’s attention towards common-quality anchor boxes to reduce the adverse effects of inconsistent annotations. The experimental results demonstrate that the improved network achieves a mean average precision (mAP) of 98.4% on the bearing dataset, representing a 4.2% improvement over the original network.

SWISTA-Nets: Subband-adaptive wavelet iterative shrinkage thresholding networks for image reconstruction

Robust and interpretable image reconstruction is central to imageology applications in clinical practice. Prevalent deep networks, with strong learning ability to extract implicit information from data manifold, are still lack of prior knowledge introduced from mathematics or physics, leading to instability, poor structure interpretability and high computation cost. As to this issue, we propose two prior knowledge-driven networks to combine the good interpretability of mathematical methods and the powerful learnability of deep learning methods. Incorporating different kinds of prior knowledge, we propose subband-adaptive wavelet iterative shrinkage thresholding networks (SWISTA-Nets), where almost every network module is in one-to-one correspondence with each step involved in the iterative algorithm. By end-to-end training of proposed SWISTA-Nets, implicit information can be extracted from training data and guide the tuning process of key parameters that possess mathematical definition. The inverse problems associated with two medical imaging modalities, i.e., electromagnetic tomography and X-ray computational tomography are applied to validate the proposed networks. Both visual and quantitative results indicate that the SWISTA-Nets outperform mathematical methods and state-of-the-art prior knowledge-driven networks, especially with fewer training parameters, interpretable network structures and well robustness. We assume that our analysis will support further investigation of prior knowledge-driven networks in the field of ill-posed image reconstruction.

TSoft-Net: A novel transfer soft thresholding network based on self-attention for intelligent fault diagnosis of rotating machinery

In the industrial scenes, the machinery operates under diverse working conditions and generates varying levels of noise, which can hinder the performance of the intelligent fault diagnosis model that is trained in the laboratory. This is because the different working conditions and environmental noise change the distribution of laboratory vibration signal. To tackle this issue, we proposed a novel transfer soft thresholding network (TSoft-Net) based on self-attention for intelligent fault diagnosis of rotating machinery, which has good anti-noise performance and can be transferred to different working conditions with excellent accuracy. This paper constructs a soft residual block for extracting the fault representation and enhancing the residual learning. In this block, we propose a residual factor to learn and enhance domain-invariant fault representation. Furthermore, a representation soft fusion block is built for extracting and fusing the different scale fault representation. In this block, we propose a scale-attention weight to dynamically fuse the different scale fault representation. The experiments show that the TSoft-Net, compared with six existing methods on eight sub-datasets, has better anti-noise ability and achieves an accuracy of up to 100% on the target domain.

Analysis of microvascular pattern in diabetes mellitus condition using the nailfold capillaroscopy images.

Diabetes is often considered a vascular disease due to its impact on blood vessels, it is a complex condition with various metabolic and autoimmune factors involved. One of the long term comorbidities of diabetes includes microvascular complications. The microvascular complications can be analyzed using the Nailfold capillaroscopy, a non-invasive technique that allows for the visualization and analysis of capillaries in the proximal nailfold area. Using advanced video capillaroscopy with high magnification, capillary images can be captured from and processed to analyze their morphology. The capillary images of normal group and diabetic group are acquired from 118 participants using nailfold capillaroscopy and the obtained images are preprocessed using image processing filters. The identification and segmentation of the capillaries are the challenges to be addressed in the processing of the images. Hence segmentation of capillaries is done using morphological operations, thresholding and convolutional neural networks. The performance of the filters and segmentation methods are evaluated using Mean Square Error (MSE), Peak signal to Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Jaccard Index and Sorensen coefficient. By analyzing the morphological features namely the capillary diameter, density, distribution, presence of hemorrhage and the shape of the capillaries from both the groups, the capillary changes associated with diabetic condition were studied. It was found that the non diabetic participants considered in this study has capillary diameter in the range of 8-14 µm and the capillary density in the range of 10-30 capillaries per mm2 whereas the diabetic participants has capillary diameter greater than 30 µm and the capillary density is less than 10 capillaries per mm2. In addition to capillary density and diameter, the presence of hemorrhage, the orientation and distribution of the capillaries are also considered to differentiate the diabetic group from the non diabetic group. The classification of the participants are validated with the clinical history of the participants.

Enhancing Exchange-Traded Fund Price Predictions: Insights from Information-Theoretic Networks and Node Embeddings.

This study presents a novel approach to predicting price fluctuations for U.S. sector index ETFs. By leveraging information-theoretic measures like mutual information and transfer entropy, we constructed threshold networks highlighting nonlinear dependencies between log returns and trading volume rate changes. We derived centrality measures and node embeddings from these networks, offering unique insights into the ETFs' dynamics. By integrating these features into gradient-boosting algorithm-based models, we significantly enhanced the predictive accuracy. Our approach offers improved forecast performance for U.S. sector index futures and adds a layer of explainability to the existing literature.

Brain MRI detection and classification: Harnessing convolutional neural networks and multi-level thresholding.

Brain tumor detection in clinical applications is a complex and challenging task due to the intricate structures of the human brain. Magnetic Resonance (MR) imaging is widely preferred for this purpose because of its ability to provide detailed images of soft brain tissues, including brain tissue, cerebrospinal fluid, and blood vessels. However, accurately detecting brain tumors from MR images remains an open problem for researchers due to the variations in tumor characteristics such as intensity, texture, size, shape, and location. To address these issues, we propose a method that combines multi-level thresholding and Convolutional Neural Networks (CNN). Initially, we enhance the contrast of brain MR images using intensity transformations, which highlight the infected regions in the images. Then, we use the suggested CNN architecture to classify the enhanced MR images into normal and abnormal categories. Finally, we employ multi-level thresholding based on Tsallis entropy (TE) and differential evolution (DE) to detect tumor region(s) from the abnormal images. To refine the results, we apply morphological operations to minimize distortions caused by thresholding. The proposed method is evaluated using the widely used Harvard Medical School (HMS) dataset, and the results demonstrate promising performance with 99.5% classification accuracy and 92.84% dice similarity coefficient. Our approach outperforms existing state-of-the-art methods in brain tumor detection and automated disease diagnosis from MR images.

A Multiscale Generalized Shrinkage Threshold Network for Image Blind Deblurring in Remote Sensing

A Multiscale Generalized Shrinkage Threshold Network for Image Blind Deblurring in Remote Sensing

Size and Depth of Monotone Neural Networks: Interpolation and Approximation.

We study monotone neural networks with threshold gates where all the weights (other than the biases) are nonnegative. We focus on the expressive power and efficiency of the representation of such networks. Our first result establishes that every monotone function over [0,1]d can be approximated within arbitrarily small additive error by a depth-4 monotone network. When , we improve upon the previous best-known construction, which has a depth of d+1 . Our proof goes by solving the monotone interpolation problem for monotone datasets using a depth-4 monotone threshold network. In our second main result, we compare size bounds between monotone and arbitrary neural networks with threshold gates. We find that there are monotone real functions that can be computed efficiently by networks with no restriction on the gates, whereas monotone networks approximating these functions need exponential size in the dimension.

Novel cross-domain fault diagnosis method based on model-agnostic meta-learning embedded in adaptive threshold network

Fault diagnosis of rotating machinery based on deep learning requires many identical labeled data to achieve satisfactory results. However, in engineering scenarios, few typical fault data have been manually marked, resulting in a poor on-site diagnosis consequence. Thus, training an efficient cross-domain diagnosis model under limited data is extremely challenging. This paper proposes a novel cross-domain fault diagnosis method based on model-agnostic meta-learning embedded in Adaptive threshold network (ATN-MAML). The threshold acquisition module is built to determine the signal's irrelevant information threshold adaptively. The designed threshold function is embedded in the network to filter out invalid information effectively as a non-linear layer. Then, MAML's training strategy is optimized to prevent overfitting. Finally, many cross-domain comparison experiments are conducted to verify the effectiveness of the proposed method on two datasets. The results show the model can maintain high accuracy even when the source and target domain data are from different devices.

LogBTF: gene regulatory network inference using Boolean threshold network model from single-cell gene expression data.

From a systematic perspective, it is crucial to infer and analyze gene regulatory network (GRN) from high-throughput single-cell RNA sequencing data. However, most existing GRN inference methods mainly focus on the network topology, only few of them consider how to explicitly describe the updated logic rules of regulation in GRNs to obtain their dynamics. Moreover, some inference methods also fail to deal with the over-fitting problem caused by the noise in time series data. In this article, we propose a novel embedded Boolean threshold network method called LogBTF, which effectively infers GRN by integrating regularized logistic regression and Boolean threshold function. First, the continuous gene expression values are converted into Boolean values and the elastic net regression model is adopted to fit the binarized time series data. Then, the estimated regression coefficients are applied to represent the unknown Boolean threshold function of the candidate Boolean threshold network as the dynamical equations. To overcome the multi-collinearity and over-fitting problems, a new and effective approach is designed to optimize the network topology by adding a perturbation design matrix to the input data and thereafter setting sufficiently small elements of the output coefficient vector to zeros. In addition, the cross-validation procedure is implemented into the Boolean threshold network model framework to strengthen the inference capability. Finally, extensive experiments on one simulated Boolean value dataset, dozens of simulation datasets, and three real single-cell RNA sequencing datasets demonstrate that the LogBTF method can infer GRNs from time series data more accurately than some other alternative methods for GRN inference. The source data and code are available at https://github.com/zpliulab/LogBTF.

Gene regulatory networks with binary weights

An evolutionary computation framework to learn binary threshold networks is presented. Inspired by the recent trend of binary neural networks, where weights and activation thresholds are represented using 1 and -1 such that they can be stored in 1-bit instead of full precision, we explore this approach for gene regulatory network modeling. We test our method by inferring binary threshold networks of two regulatory network models: Quorum sensing systems in bacterium Paraburkholderia phytofirmans PsJN and the fission yeast cell-cycle. We considered differential evolution and particle swarm optimization for the simulations. Results for weights having only 1 and -1 values, and different activation thresholds are presented. Full binary threshold networks were found with minimum error (2 bits), whereas when the binary restriction is relaxed for the activation thresholds, networks with 0 bit error were found.

Analysis of Temporal Robustness in Massive Machine Type Communications

The evolution of fifth generation (5G) networks needs to support the latest use cases, which demand robust network connectivity for the collaborative performance of the network agents, like multi-robot systems and vehicle to anything (V2X) communication. Unfortunately, the user device’s limited communication range and battery constraint confirm the unfitness of known robustness metrics suggested for fixed networks, when applied to time-switching communication graphs. Furthermore, the calculation of most of the existing robustness metrics involves non-deterministic polynomial (NP)-time complexity, and hence are best-fitted only for small networks. Despite a large volume of works, the complete analysis of a low-complexity temporal robustness metric for a communication network is absent in the literature, and the present work aims to fill this gap. More in detail, our work provides a stochastic analysis of network robustness for a massive machine type communication (mMTC) network. The numerical investigation corroborates the exactness of the proposed analytical framework for temporal robustness metric. Along with studying the impact on network robustness of various system parameters, such as cluster head (CH) probability, power threshold value, network size, and node failure probability, we justify the observed trend of numerical results probabilistically.

Predicting Low-Level Childhood Lead Exposure in Metro Atlanta Using Ensemble Machine Learning of High-Resolution Raster Cells.

Low-level lead exposure in children is a major public health issue. Higher-resolution spatial targeting would significantly improve county and state-wide policies and programs for lead exposure prevention that generally intervene across large geographic areas. We use stack-ensemble machine learning, including an elastic net generalized linear model, gradient-boosted machine, and deep neural network, to predict the number of children with venous blood lead levels (BLLs) ≥2 to <5 µg/dL and ≥5 µg/dL in ~1 km2 raster cells in the metro Atlanta region using a sample of 92,792 children ≤5 years old screened between 2010 and 2018. Permutation-based predictor importance and partial dependence plots were used for interpretation. Maps of predicted vs. observed values were generated to compare model performance. According to the EPA Toxic Release Inventory for air-based toxic release facility density, the percentage of the population below the poverty threshold, crime, and road network density was positively associated with the number of children with low-level lead exposure, whereas the percentage of the white population was inversely associated. While predictions generally matched observed values, cells with high counts of lead exposure were underestimated. High-resolution geographic prediction of lead-exposed children using ensemble machine learning is a promising approach to enhance lead prevention efforts.