Search

A neural network was developed to detect and characterize bowel obstruction, a common cause of acute abdominal pain. In this retrospective study, 202 CT scans of 165 patients with bowel obstruction from March to June 2022 were included and partitioned into training and test data sets. A multi-channel neural network was trained to segment the gastrointestinal tract, and to predict the diameter and the longitudinal position (longitude) along the gastrointestinal tract using a novel embedding. Its performance was compared to manual segmentations using the Dice score, and to manual measurements of the diameter and longitude using intraclass correlation coefficients (ICC). ROC curves as well as sensitivity and specificity were calculated for diameters above a clinical threshold for obstruction, and for longitudes corresponding to small bowel. In the test data set, Dice score for segmentation of the gastrointestinal tract was 78 ± 8%. ICC between measured and predicted diameters was 0.72, indicating moderate agreement. ICC between measured and predicted longitude was 0.85, indicating good agreement. AUROC was 0.90 for detection of dilated bowel, and was 0.95 and 0.90 for differentiation of the proximal and distal gastrointestinal tract respectively. Overall sensitivity and specificity for dilated small bowel were 0.83 and 0.90. Since obstruction is diagnosed based on the diameter and longitude of the bowel, this neural network and embedding may enable detection and characterization of this important disease on CT.

Copyright of Paint & Coatings Industry (0253-4312) is the property of Paint & Coatings Industry Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Continue Education Research is the property of Harbin Normal University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

We consider k-dimensional discrete-time systems of the form x n + 1 = F (x n , … , x n - k + 1) in which the map F is continuous and monotonic in each one of its arguments. We define a partial order on R + 2 k , compatible with the monotonicity of F, and then use it to embed the k-dimensional system into a 2k-dimensional system that is monotonic with respect to this poset structure. An analogous construction is given for periodic systems. Using the characteristics of the higher-dimensional monotonic system, global stability results are obtained for the original system. Our results apply to a large class of difference equations that are pertinent in a variety of contexts. As an application of the developed theory, we provide two examples that cover a wide class of difference equations, and in a subsequent paper, we provide additional applications of general interest. [ABSTRACT FROM AUTHOR]

Data-driven machine learning models are useful for modeling complex-typically nonlinear-structures based on empirical observations, bypassing the need to generate a physical model in cases where the physics is not well known or easily modeled. One disadvantage of purely data-driven approaches is that they tend to perform poorly in regions outside the original training domain. To mitigate this limitation, physical knowledge about the structure can be embedded in the model architecture via the model topology or numerical constraints in the formulation. We propose a neural network framework based on Hamiltonian mechanics to enforce a physics-informed structure to the model. The Hamiltonian framework allows us to relate the energy of the system to the measured quantities (e.g., accelerations) through the Euler–Lagrange equations of motion. A challenge with this hybrid data-driven, physics-constrained approach is the problem of limited observability, i.e., not being able to measure structural response in a complete coordinate system that is compatible with the physical constraints being enforced. To overcome this issue, we propose combining an invertible neural network autoencoder architecture with knowledge from embedding theory to enrich the limited observable data with time-delay embeddings. From Taken's theorem, we know that a sufficient time-delay embedding is diffeomorphically equivalent to the underlying state space of the system. We use this information to find time-delays of the original data and build the diffeomorphic mapping with a neural network encoder. The approach is demonstrated on computational and experimental examples. [ABSTRACT FROM AUTHOR]

Traffic sensors are vital to the development and operation of Intelligent Transportation Systems, providing essential data for traffic monitoring, management, and transportation infrastructure planning. However, optimizing the placement of these sensors, particularly across large and complex statewide highway networks, remains a challenging task. In this research, we presented a novel search algorithm designed to address this challenge by leveraging information gradients from K-nearest neighbors within an embedding space. Our method enabled more informed and strategic sensor placement under budget and resource constraints, enhancing overall network coverage and data quality. Additionally, we incorporated spatial kriging analysis, harnessing spatial correlations of existing sensors to refine and reduce the search space. Our proposed approach was tested against the widely used Genetic Algorithm, demonstrating superior efficiency in terms of convergence time and producing more effective solutions with reduced information loss. [ABSTRACT FROM AUTHOR]

Understanding the intricacies of plant biology demands innovative analytical methods to elucidate genetic and molecular mechanisms underlying plant development and stress responses. Large language models (LLMs) can find patterns and correlations in noisy biological data, but LLMs have not yet been embraced by the plant community. We describe the different types of LLMs and cover how they can be used to study biological systems. The data, hardware, and software requirements are within reach to build powerful LLMs for the plant community. Large language models (LLMs), such as ChatGPT, have taken the world by storm. However, LLMs are not limited to human language and can be used to analyze sequential data, such as DNA, protein, and gene expression. The resulting foundation models can be repurposed to identify the complex patterns within the data, resulting in powerful, multipurpose prediction tools able to predict the state of cellular systems. This review outlines the different types of LLMs and showcases their recent uses in biology. Since LLMs have not yet been embraced by the plant community, we also cover how these models can be deployed for the plant kingdom. [ABSTRACT FROM AUTHOR]

We introduce a new embedding technique based on a nested barycentric coordinate system. We show that our embedding can be used to transform the problems of polyhedron approximation, piecewise linear classification and convex regression into one of finding a linear classifier or regressor in a higher dimensional (but nevertheless quite sparse) representation. Our embedding maps a piecewise linear function into an everywhere-linear function, and allows us to invoke well-known algorithms for the latter problem to solve the former. We explain the applications of our embedding to the problems of approximating separating polyhedra—in fact, it can approximate any convex body and unions of convex bodies—as well as to classification by separating polyhedra, and to piecewise linear regression. [ABSTRACT FROM AUTHOR]

We consider the Ricker model with delay and constant or periodic stocking. We found that the high stocking density tends to neutralize the delay effect on stability. Conditions are established on the parameters to ensure the global stability of the equilibrium solution in the case of constant stocking, as well as the global stability of the 2‐periodic solution in the case of 2‐periodic stocking. Our approach extensively relies on the utilization of the embedding technique. Whether constant stocking or periodic stocking, the model has the potential to undergo a Neimark–Sacker bifurcation in both cases. However, the Neimark–Sacker bifurcation in the 2‐periodic case results in the emergence of two invariant curves that collectively function as a single attractor. Finally, we pose open questions in the form of conjectures about global stability for certain choices of the parameters. [ABSTRACT FROM AUTHOR]

Technological advances in information-processing capacity have enabled integrated analyses (multi-omics) of different omics data types, improving target discovery and clinical diagnosis. This study proposes novel artificial intelligence (AI) learning strategies for incomplete datasets, common in omics research. The model comprises (1) a multi-omics generative model based on a variational auto-encoder that learns tumor genetic patterns based on different omics data types and (2) an expanded classification model that predicts cancer phenotypes. Padding was applied to replace missing data with virtual data. The embedding data generated by the model accurately classified cancer phenotypes, addressing the class imbalance issue (weighted F1 score: cancer type > 0.95, primary site > 0.92, sample type > 0.97). The classification performance was maintained in the absence of omics data, and the virtual data resembled actual omics data (cosine similarity mRNA gene expression > 0.96, mRNA isoform expression > 0.95, DNA methylation > 0.96). Meanwhile, in the presence of omics data, high-quality, non-existent omics data were generated (cosine similarity mRNA gene expression: 0.9702, mRNA isoform expression: 0.9546, DNA methylation: 0.9687). This model can effectively classify cancer phenotypes based on incomplete omics data with data sparsity robustness, generating omics data through deep learning and enabling precision medicine. [ABSTRACT FROM AUTHOR]

In digital media copyright protection, image watermarking topologies afford a promising solution. But the robustness of watermarking methods should be considered. Therefore, a robust image watermarking technique has been proposed to show better robustness against rotation attacks and other issues in watermarking. In this work, the Deep Belief Network (DBN) is trained by the Bear Smell Search Algorithm (BSSA) during the first phase. The second phase is the embedding phase, which utilizes the hybrid transform domain of Discrete Wavelet Transform and Singular Value Decomposition (DWT-SVD). After embedding, the extraction is performed with the help of a Back Propagation Neural Network (BPNN). The experiments are conducted on four kinds of host images such as Barbara, Peppers, Person and Lena. Experimental results proved that the superiority of our proposed method achieving the state-of-the-art accuracy. [ABSTRACT FROM AUTHOR]

Technological advances in information-processing capacity have enabled integrated analyses (multi-omics) of different omics data types, improving target discovery and clinical diagnosis. This study proposes novel artificial intelligence (AI) learning strategies for incomplete datasets, common in omics research. The model comprises (1) a multi-omics generative model based on a variational auto-encoder that learns tumor genetic patterns based on different omics data types and (2) an expanded classification model that predicts cancer phenotypes. Padding was applied to replace missing data with virtual data. The embedding data generated by the model accurately classified cancer phenotypes, addressing the class imbalance issue (weighted F1 score: cancer type > 0.95, primary site > 0.92, sample type > 0.97). The classification performance was maintained in the absence of omics data, and the virtual data resembled actual omics data (cosine similarity mRNA gene expression > 0.96, mRNA isoform expression > 0.95, DNA methylation > 0.96). Meanwhile, in the presence of omics data, high-quality, non-existent omics data were generated (cosine similarity mRNA gene expression: 0.9702, mRNA isoform expression: 0.9546, DNA methylation: 0.9687). This model can effectively classify cancer phenotypes based on incomplete omics data with data sparsity robustness, generating omics data through deep learning and enabling precision medicine.

We consider the problem of finding an accurate representation of neuron shapes, extracting sub-cellular features, and classifying neurons based on neuron shapes. In neuroscience research, the skeleton representation is often used as a compact and abstract representation of neuron shapes. However, existing methods are limited to getting and analyzing curve skeletons which can only be applied for tubular shapes. This paper presents a 3D neuron morphology analysis method for more general and complex neuron shapes. First, we introduce the concept of skeleton mesh to represent general neuron shapes and propose a novel method for computing mesh representations from 3D surface point clouds. A skeleton graph is then obtained from skeleton mesh and is used to extract sub-cellular features. Finally, an unsupervised learning method is used to embed the skeleton graph for neuron classification. Extensive experiment results are provided and demonstrate the robustness of our method to analyze neuron morphology.

An important tool for the execution of parallel algorithms and the simulation of interconnection networks is graph embedding. The quality of an embedding can be assessed using some cost metrics. The dilation and wirelength are the commonly used parameters. The Knödel graph W Δ , n is a minimum linear gossip network and has minimum broadcasting. It has n vertices, n Δ 2 edges, where n is even, and 1 ≤ Δ ≤ ⌊ log 2   n ⌋. In this study, we solve the dilation problem of embedding the Knödel graph into certain cube-like architectures such as hypercube, folded hypercube, and augmented cube. In [G. Fertin, A. Raspaud, A survey on Knödel graphs, Discrete Applied Mathematics 137 (2004) 173–195], it is proved that the dilation of embedding the Knödel graph W Δ , 2 Δ into the hypercube Q Δ is at most 4. In this study, we obtain an improved upper bound for dilation of embedding the Knödel graph into the hypercube and it is equal to 3. Also, we calculate the wirelength of embedding the Knödel graph into the above-said cube-like architectures using dilation. [ABSTRACT FROM AUTHOR]

Finding the source of rumors in the social network was addressed by researchers with probabilistic models like Maximum Likelihood Estimation in complex network analysis for the past few decades. However, the most promising results could reach up to 2-hop distant neighborhoods on average. With the advent of graph neural networks, the issue was addressed as a classical node classification problem in large networks. Node classification problems achieve the best results when there are appropriate node attributes as features and when node classes are balanced. However, unlike other node classification scenarios, the data collected for source identification usually lacks node attributes because of time limitations. Moreover, the detection of the sources among thousands of users is typically a class imbalance problem. If we could deal with these issues skillfully, then the dominance of the deep learning method compared to other conventional probabilistic methods in multiple rumor source detection will be prominent. We have proposed here a deep learning-based multiple source node classification framework that can predict the sources with promising accuracy. The primary hurdles in non-attributed network classification are navigated by generating feature vectors that capture the structural characteristics of the network and spreading pattern. These features are further solidified with the Graph embedding technique, incorporating the neighborhood features. We have triumphed over the challenge of imbalanced node classes by synthetic node generation with a suitable mathematical model. The concern is further resolved by selective sampling and weighted loss estimation in the deep learning network for classification used in the framework. A study of the Immuno-Diffuse Likelihood parameter in Label propagation based feature construction and its influence on accurate prediction is examined. Our approach demonstrates superior performance compared to existing methods in the datasets available in public repositories, making it a reliable and robust tool for rumor source detection in the complex landscape of social media. [ABSTRACT FROM AUTHOR]

Analysis of the impact of the contextual embeddings usage on the text classification accuracy

Trichloroacetic acid is a common nonvolatile byproduct of drinking water disinfection and poses carcinogenic risks to the human body. In this study, four types of nano-oxide@microcrystalline-cellulose-based adsorbents (P25@microcrystalline cellulose, SiO₂@microcrystalline cellulose (MCC-g-GMA@SiO2), Fe3O4@microcrystalline cellulose, and Fe2O3@microcrystalline cellulose) were successfully prepared by the pre-radiation grafting-embedding method. Subsequently, their ability to remove trichloroacetic acid from drinking water was investigated. The micro zonation morphology and surface properties of the materials were characterized and tested using Fourier transform infrared spectroscopy (FTIR), X-ray diffraction spectroscopy (XRD), Thermogravimetric (TG) analysis, and Scanning electron microscopy (SEM) characterization methods. The effects of monomer concentration, embedding concentration, and absorbed dose on the weight gain rate were systematically investigated. A complete static adsorption equilibrium curve was obtained on the basis of the results of adsorption experiments of four buried nano-oxides. The performance of SiO2@microcrystalline cellulose was found to be significantly higher than that of the other three adsorbents. When the volum percentage of monomer concentration was 30%, the mass percentage of embedding concentration was 4%, and the absorbed dose was 60 kGy, the removal rate of trichloroacetic acid in drinking water reached 83.27%. This series of adsorbent materials present significant potential for practical application in drinking water purification.

Abstract Background Electronic Health Records (EHR) are widely used to develop clinical prediction models (CPMs). However, one of the challenges is that there is often a degree of informative missing data. For example, laboratory measures are typically taken when a clinician is concerned that there is a need. When data are the so-called Not Missing at Random (NMAR), analytic strategies based on other missingness mechanisms are inappropriate. In this work, we seek to compare the impact of different strategies for handling missing data on CPMs performance. Methods We considered a predictive model for rapid inpatient deterioration as an exemplar implementation. This model incorporated twelve laboratory measures with varying levels of missingness. Five labs had missingness rate levels around 50%, and the other seven had missingness levels around 90%. We included them based on the belief that their missingness status can be highly informational for the prediction. In our study, we explicitly compared the various missing data strategies: mean imputation, normal-value imputation, conditional imputation, categorical encoding, and missingness embeddings. Some of these were also combined with the last observation carried forward (LOCF). We implemented logistic LASSO regression, multilayer perceptron (MLP), and long short-term memory (LSTM) models as the downstream classifiers. We compared the AUROC of testing data and used bootstrapping to construct 95% confidence intervals. Results We had 105,198 inpatient encounters, with 4.7% having experienced the deterioration outcome of interest. LSTM models generally outperformed other cross-sectional models, where embedding approaches and categorical encoding yielded the best results. For the cross-sectional models, normal-value imputation with LOCF generated the best results. Conclusion Strategies that accounted for the possibility of NMAR missing data yielded better model performance than those did not. The embedding method had an advantage as it did not require prior clinical knowledge. Using LOCF could enhance the performance of cross-sectional models but have countereffects in LSTM models.

Abstract Background Some clients who access healthcare services experience problems due to the wider determinants of health which cannot be addressed (solely) by the medical sector. Social Prescribing (SP) addresses clients ’ wider health needs and is based on linkworkers who support primary care clients in accessing social, community and voluntary care services that support their needs. Previous literature has provided valuable insights about what works (or not) in an early stage of implementing SP. However, there is limited insight into what works for the implementation of SP towards embedding. This study provides guiding principles by which SP can be successfully implemented towards the embedding stage and identifies which contextual factors and mechanisms influence these guiding principles. Methods A Rapid Realist Review was conducted to examine what works, for whom, why, and in which contexts. A local Dutch reference panel consisting of health and care organisations helped to inform the research questions. Additionally, a workshop was held with the panel, to discuss what the international insights mean for their local contexts. This input helped to further refine the literature review’s findings. Results Five guiding principles were identified for successful implementation of SP at the embedding stage: • Create awareness for addressing the wider determinants of health and the role SP services can play; • Ensure health and care professionals build trusting relationships with all involved stakeholders to create a cyclical referral process; • Invest in linkworkers’ skills and capacity so that they can act as a bridge between the sectors; • Ensure clients receive appropriate support to improve their self-reliance and increase their community participation; • Invest in the aligning of structures, processes and resources between involved sectors to support the use of SP services. Conclusion To embed SP, structural changes on a system level as well as cultural changes are needed. This will require a shift in attitude amongst health and care professionals as well as clients towards the use, role and benefit of SP services in addressing the wider determinants of health. It will also require policymakers and researchers to involve communities and include their perspectives.

Background Faced with the increasingly severe trend of population aging and the threat of non communicable chronic diseases, China actively guides and promotes the development of community embedded integrated medical and nursing care service with characteristics of traditional Chinese medicine (TCM). However, there are limited studies on performance evaluation in this field in the existing literature. The construction of a set of scientific and effective performance evaluation index system for community embedded integrated medical and nursing care service with characteristics of TCM has important theoretical significance and practical value. Objective To construct a performance evaluation index system for community embedded integrated medical and nursing care service with characteristics of TCM, determine the weights for indicators of each level, so as to provide reference for the formulation of relevant policies and the performance evaluation of community embedded integrated medical and nursing care service with characteristics of TCM. Methods From September to October 2022, a two-dimensional theoretical framework of "embedded-performance" for the community embedded integrated medical and nursing care service with characteristics of TCM was established through literature research and based on the embeddedness theory and the structure-process-outcome theoretical model. From October to November 2022, an index system was constructed based on the two-dimensional theoretical framework of "embedded-performance", with methods of literature research, Delphi method, and field investigation. In December 2022, the analytic hierarchy process was applied to determine the weight coefficients and test the logical consistency of indicators at all levels. Results A performance evaluation index system for community embedded integrated medical and nursing care service with characteristics of TCM, containing 2 dimensions of embeddedness and embedded performance, 6 primary indicators, 19 secondary indicators, and 45 tertiary indicators was constructed. The primary indicators with the highest weight were A5 (service process) and A6 (service outcome), while the secondary and tertiary indicators with the highest combined weight were B17 (service output), B19 (service satisfaction), and C44 (elderly satisfaction). All judgment matrices have passed the consistency test with CR

Background: There has been a considerable advancement in AI technologies like LLM and machine learning to support biomedical knowledge discovery. Main body: We propose a novel biomedical neural search service called 'VAIV Bio-Discovery', which supports enhanced knowledge discovery and document search on unstructured text such as PubMed. It mainly handles with information related to chemical compound/drugs, gene/proteins, diseases, and their interactions (chemical compounds/drugs-proteins/gene including drugs-targets, drug-drug, and drug-disease). To provide comprehensive knowledge, the system offers four search options: basic search, entity and interaction search, and natural language search. We employ T5slim_dec, which adapts the autoregressive generation task of the T5 (text-to-text transfer transformer) to the interaction extraction task by removing the self-attention layer in the decoder block. It also assists in interpreting research findings by summarizing the retrieved search results for a given natural language query with Retrieval Augmented Generation (RAG). The search engine is built with a hybrid method that combines neural search with the probabilistic search, BM25. Conclusion: As a result, our system can better understand the context, semantics and relationships between terms within the document, enhancing search accuracy. This research contributes to the rapidly evolving biomedical field by introducing a new service to access and discover relevant knowledge. [ABSTRACT FROM AUTHOR]

AbstractIn this paper we make explicit an application of the wreath product construction to the Galois groups of field extensions. More precisely, given a tower of fields F⊆K⊆L with L/F finite and separable, we explicitly construct an embedding of the Galois group Gal(Lc/F) into the regular wreath product Gal(Lc/Kc)≀rGal(Kc/F) . Here Lc (resp. Kc ) denotes the Galois closure of L/F (resp. K/F). Similarly, we also construct an explicit embedding of the Galois group Gal(Lc/F) into the smaller sized wreath product Gal(Lc/K)≀ΩGal(Kc/F) , where Ω=HomF(K,Kc) is acted on by composition of automorphisms in Gal(Kc/F) . Moreover, when L/K is a Kummer extension we prove a sharper embedding, that is, that Gal(Lc/F) embeds into the wreath product Gal(L/K)≀ΩGal(Kc/F) . As corollaries we obtain embedding theorems when L/K is cyclic and when it is quadratic with char(F)≠2 . We also provide examples of these embeddings and as an illustration of the usefulness of these embedding theorems, we survey some recent applications of these types of results in field theory, arithmetic statistics, number theory and arithmetic geometry. [ABSTRACT FROM AUTHOR]

We present an algorithm determining the existence of an isomorphic embedding of one finitely generated partially commutative nilpotent group of class in another. It is shown how such embeddings can be constructed. We also describe an algorithm determining the existence of an embedding of a finitely generated partially commutative nilpotent group of arbitrary class with respect to a graph of nonzero radius in a free nilpotent group of class . [ABSTRACT FROM AUTHOR]

Every countable group G can be embedded in a finitely generated group G* that is hopfian and complete, that is, G* has trivial centre and every epimorphism G* → G* is an inner automorphism. Every finite subgroup of G* is conjugate to a finite subgroup of G. If G has a finite presentation (respectively, a finite classifying space), then so does G*. Our construction of G* relies on the existence of closed hyperbolic 3-manifolds that are asymmetric and non-Haken. [ABSTRACT FROM AUTHOR]

This essay argues that Iraq war reportage participates in an embedded sublime that offers aestheticized and racialized spectacles of war at a distance. The US military media embedding program aimed to decisively shape the media's representation of the 2003 invasion and occupation of Iraq. But embedding journalists within military units proved challenging, as the programme confronted both the ironies of framed narratives and a destabilising media and internet environment. Reading Anthony Swofford's Jarhead, Evan Wright's Generation Kill, and Dexter Filkins' The Forever War, I argue that their hypermasculine representations of war show a powerful desire for poeticized violence and for scenes of despicable beauty. These scenes draw on classical formulations of the sublime in Kant, depending on racial hierarchies and imperial narratives as they represent the killing and suffering of people of colour in Iraq, Afghanistan and the Global South. Yet the embedded sublime often produces an ethical hesitation in readers and viewers allowing them to participate as spectators of war at a distance. I explore strategies that challenge this logic in Iraqi blogger Riverbend's Baghdad Burning and journalist Nick McDonell's The End of Major Combat Operations, resisting this tendency to embed, aestheticize and normalise representations of violence in the 'The Forever War' era. [ABSTRACT FROM AUTHOR]

The performance of the recently suggested projected atomic orbitals (PAO) approach as the virtual space in excited state projection-based embedding applications is analysed. The role of the parameters in the PAO generation is discussed and the impact of different choices is evaluated on the ground state of the stacked formaldehyde and pyrrole dimers. A comparison of excitation energies obtained with PAOs and localised virtual orbitals is given and the role of diffuse basis functions is discussed using a benchmark set from previous studies. [ABSTRACT FROM AUTHOR]

Graph-based semi-supervised learning plays an important role in large scale image classification tasks. However, the problem becomes very challenging in the presence of noisy labels and outliers. Moreover, traditional robust semi-supervised learning solutions suffers from prohibitive computational burdens thus cannot be computed for streaming data. Motivated by that, we present a novel unified framework robust structure-aware semi-supervised learning called Unified RSSL (URSSL) for batch processing and recursive processing robust to both outliers and noisy labels. Particularly, URSSL applies joint semi-supervised dimensionality reduction with robust estimators and network sparse regularization simultaneously on the graph Laplacian matrix iteratively to preserve the intrinsic graph structure and ensure robustness to the compound noise. First, in order to relieve the influence from outliers, a novel semi-supervised robust dimensionality reduction is applied relying on robust estimators to suppress outliers. Meanwhile, to tackle noisy labels, the denoised graph similarity information is encoded into the network regularization. Moreover, by identifying strong relevance of dimensionality reduction and network regularization in the context of robust semi-supervised learning (RSSL), a two-step alternative optimization is derived to compute optimal solutions with guaranteed convergence. We further derive our framework to adapt to large scale semi-supervised learning particularly suitable for large scale image classification and demonstrate the model robustness under different adversarial attacks. For recursive processing, we rely on reparameterization to transform the formulation to unlock the challenging problem of robust streaming-based semi-supervised learning. Last but not least, we extend our solution into distributed solutions to resolve the challenging issue of distributed robust semi-supervised learning when images are captured by multiple cameras at different locations. Extensive experimental results demonstrate the promising performance of this framework when applied to multiple benchmark datasets with respect to state-of-the-art approaches for important applications in the areas of image classification and spam data analysis. [ABSTRACT FROM AUTHOR]

Göring–Helmberg–Wappler introduced optimization problems regarding embeddings of a graph into a Euclidean space and the first nonzero eigenvalue of the Laplacian of a graph, which are dual to each other in the framework of semidefinite programming. In this paper, we introduce a new graph-embedding optimization problem, and discuss its relation to Göring–Helmberg–Wappler's problems. We also identify the dual problem to our embedding optimization problem. We solve the optimization problems for distance-regular graphs and the one-skeleton graphs of the C 60 fullerene and some other Archimedian solids. [ABSTRACT FROM AUTHOR]

Accurately modeling the effects of illumination and shadows during head rotation is critical in computer vision for enhancing image realism and reducing artifacts. This study delves into the latent space of denoising diffusion models to identify compelling trajectories that can express continuous head rotation under varying lighting conditions. A key contribution of our work is the generation of additional labels from the CelebA dataset, categorizing images into three groups based on prevalent illumination direction: left, center, and right. These labels play a crucial role in our approach, enabling more precise manipulations and improved handling of lighting variations. Leveraging a recent embedding technique for Denoising Diffusion Implicit Models (DDIM), our method achieves noteworthy manipulations, encompassing a wide rotation angle of ±30°. while preserving individual distinct characteristics even under challenging illumination conditions. Our methodology involves computing trajectories that approximate clouds of latent representations of dataset samples with different yaw rotations through linear regression. Specific trajectories are obtained by analyzing subsets of data that share significant attributes with the source image, including light direction. Notably, our approach does not require any specific training of the generative model for the task of rotation; we merely compute and follow specific trajectories in the latent space of a pre-trained face generation model. This article showcases the potential of our approach and its current limitations through a qualitative discussion of notable examples. This study contributes to the ongoing advancements in representation learning and the semantic investigation of the latent space of generative models. [ABSTRACT FROM AUTHOR]

A neural network was developed to detect and characterize bowel obstruction, a common cause of acute abdominal pain. In this retrospective study, 202 CT scans of 165 patients with bowel obstruction from March to June 2022 were included and partitioned into training and test data sets. A multi-channel neural network was trained to segment the gastrointestinal tract, and to predict the diameter and the longitudinal position ("longitude") along the gastrointestinal tract using a novel embedding. Its performance was compared to manual segmentations using the Dice score, and to manual measurements of the diameter and longitude using intraclass correlation coefficients (ICC). ROC curves as well as sensitivity and specificity were calculated for diameters above a clinical threshold for obstruction, and for longitudes corresponding to small bowel. In the test data set, Dice score for segmentation of the gastrointestinal tract was 78 ± 8%. ICC between measured and predicted diameters was 0.72, indicating moderate agreement. ICC between measured and predicted longitude was 0.85, indicating good agreement. AUROC was 0.90 for detection of dilated bowel, and was 0.95 and 0.90 for differentiation of the proximal and distal gastrointestinal tract respectively. Overall sensitivity and specificity for dilated small bowel were 0.83 and 0.90. Since obstruction is diagnosed based on the diameter and longitude of the bowel, this neural network and embedding may enable detection and characterization of this important disease on CT. [ABSTRACT FROM AUTHOR]

Background: Electronic Health Records (EHR) are widely used to develop clinical prediction models (CPMs). However, one of the challenges is that there is often a degree of informative missing data. For example, laboratory measures are typically taken when a clinician is concerned that there is a need. When data are the so-called Not Missing at Random (NMAR), analytic strategies based on other missingness mechanisms are inappropriate. In this work, we seek to compare the impact of different strategies for handling missing data on CPMs performance. Methods: We considered a predictive model for rapid inpatient deterioration as an exemplar implementation. This model incorporated twelve laboratory measures with varying levels of missingness. Five labs had missingness rate levels around 50%, and the other seven had missingness levels around 90%. We included them based on the belief that their missingness status can be highly informational for the prediction. In our study, we explicitly compared the various missing data strategies: mean imputation, normal-value imputation, conditional imputation, categorical encoding, and missingness embeddings. Some of these were also combined with the last observation carried forward (LOCF). We implemented logistic LASSO regression, multilayer perceptron (MLP), and long short-term memory (LSTM) models as the downstream classifiers. We compared the AUROC of testing data and used bootstrapping to construct 95% confidence intervals. Results: We had 105,198 inpatient encounters, with 4.7% having experienced the deterioration outcome of interest. LSTM models generally outperformed other cross-sectional models, where embedding approaches and categorical encoding yielded the best results. For the cross-sectional models, normal-value imputation with LOCF generated the best results. Conclusion: Strategies that accounted for the possibility of NMAR missing data yielded better model performance than those did not. The embedding method had an advantage as it did not require prior clinical knowledge. Using LOCF could enhance the performance of cross-sectional models but have countereffects in LSTM models. [ABSTRACT FROM AUTHOR]

Background: Some clients who access healthcare services experience problems due to the wider determinants of health which cannot be addressed (solely) by the medical sector. Social Prescribing (SP) addresses clients ' wider health needs and is based on linkworkers who support primary care clients in accessing social, community and voluntary care services that support their needs. Previous literature has provided valuable insights about what works (or not) in an early stage of implementing SP. However, there is limited insight into what works for the implementation of SP towards embedding. This study provides guiding principles by which SP can be successfully implemented towards the embedding stage and identifies which contextual factors and mechanisms influence these guiding principles. Methods: A Rapid Realist Review was conducted to examine what works, for whom, why, and in which contexts. A local Dutch reference panel consisting of health and care organisations helped to inform the research questions. Additionally, a workshop was held with the panel, to discuss what the international insights mean for their local contexts. This input helped to further refine the literature review's findings. Results: Five guiding principles were identified for successful implementation of SP at the embedding stage: • Create awareness for addressing the wider determinants of health and the role SP services can play; • Ensure health and care professionals build trusting relationships with all involved stakeholders to create a cyclical referral process; • Invest in linkworkers' skills and capacity so that they can act as a bridge between the sectors; • Ensure clients receive appropriate support to improve their self-reliance and increase their community participation; • Invest in the aligning of structures, processes and resources between involved sectors to support the use of SP services. Conclusion: To embed SP, structural changes on a system level as well as cultural changes are needed. This will require a shift in attitude amongst health and care professionals as well as clients towards the use, role and benefit of SP services in addressing the wider determinants of health. It will also require policymakers and researchers to involve communities and include their perspectives. [ABSTRACT FROM AUTHOR]

For an integer r ≥ 2, the space of r-immersions of M in ℝn is defined to be the space of immersions of M in ℝn such that at most r -- 1 points of M are mapped to the same point in ℝn. The space of r-immersions lies "between" the embeddings and the immersions. We calculate the connectivity of the layers in the homological Taylor tower for the space of r-immersions in ℝn (modulo immersions), and give conditions that guarantee that the connectivity of the maps in the tower approaches infinity as one goes up the tower. We also compare the homological tower with the homotopical tower, and show that up to degree 2r -- 1 there is a "Hurewicz isomorphism" between the first non-trivial homotopy groups of the layers of the two towers. [ABSTRACT FROM AUTHOR]

The digital revolution places great emphasis on digital media watermarking due to the increased vulnerability of multimedia content to unauthorized alterations. Recently, in the digital boom in the technology of hiding data, research has been tending to perform watermarking with numerous architectures of deep learning, which has explored a variety of problems since its inception. Several watermarking approaches based on deep learning have been proposed, and they have proven their efficiency compared to traditional methods. This paper summarizes recent developments in conventional and deep learning image and video watermarking techniques. It shows that although there are many conventional techniques focused on video watermarking, there are yet to be any deep learning models focusing on this area; however, for image watermarking, different deep learning-based techniques where efficiency in invisibility and robustness depends on the used network architecture are observed. This study has been concluded by discussing possible research directions in deep learning-based video watermarking. [ABSTRACT FROM AUTHOR]

This paper introduces an effective blind watermarking algorithm for double-color images utilizing the Walsh–Hadamard Transform (WHT) with symmetric embedding locations to enhance imperceptibility. The proposed algorithm leverages the energy accumulation capability and significant correlations among coefficients of the WHT. First, the color host image undergoes partitioning into its respective red (R), green (G), and blue (B) channels, followed by further subdivision into 4 × 4 blocks. Through research, the algorithm determines which WHT coefficients are least visually sensitive to embedding a color image, and as a result, optimizes the embedding locations to achieve better imperceptibility. The extensive simulation results verify the superior performance of the proposed algorithm compared to other related approaches, showcasing its excellence not only in imperceptibility but also in embedding capacity and robustness. [ABSTRACT FROM AUTHOR]

For an integer r ≥ 2, the space of r-immersions of M in ℝn is defined to be the space of immersions of M in ℝn such that at most r -- 1 points of M are mapped to the same point in ℝn. The space of r-immersions lies "between" the embeddings and the immersions. We calculate the connectivity of the layers in the homological Taylor tower for the space of r-immersions in ℝn (modulo immersions), and give conditions that guarantee that the connectivity of the maps in the tower approaches infinity as one goes up the tower. We also compare the homological tower with the homotopical tower, and show that up to degree 2r -- 1 there is a "Hurewicz isomorphism" between the first non-trivial homotopy groups of the layers of the two towers. [ABSTRACT FROM AUTHOR]

The exponential expansion of information has made text feature extraction based on simple semantic information insufficient for the multidimensional recognition of textual data. In this study, we construct a text semantic structure graph based on various perspectives and introduce weight coefficients and node clustering coefficients of co-occurrence granularity to enhance the link prediction model, in order to comprehensively capture the structural information of the text. Firstly, we jointly build the semantic structure graph based on three proposed perspectives (i.e., scene semantics, text weight, and graph structure), and propose a candidate keyword set in conjunction with an information probability retrieval model. Subsequently, we propose weight coefficients of co-occurrence granularity and node clustering coefficients to improve the link prediction model based on the semantic structure graph, enabling a more comprehensive acquisition of textual structural information. Experimental results demonstrate that our research method can reveal potential correlations and obtain more complete semantic structure information, while the WPAA evaluation index validates the effectiveness of our model. [ABSTRACT FROM AUTHOR]

Drug treatment strategies to reduce dose-related hazards is a tried-and-true method for preventing drug resistance and enhancing the efficiency of the monotherapy. Except when certain drugs pile up. Most adverse medication effects are induced by antagonistic drug-drug interactions. New medications and monitoring patients' use of more effective medication combination therapies require precise Drug-Drug Interaction (DDI) prediction. Several machine learning-based DDI prediction methods exist. This wide range of strategies uses drug-related and substancerelated traits covertly. Graph embeddings and deep learning are applied to benchmark datasets to overcome this. The Simplified Molecular Input Line Entry System (SMILE) method is introduced for preprocessing, and the GCNet is applied for DDI prediction. Moreover, the graph is also constructed based on that the similarity is identified using link prediction. The proposed method provides an accuracy range of 0.934, Mean Squared Error (MSE) of 0.082, and Root Mean Squared Error (RMSE) of 0.352, which assists in more effectively reducing adverse drug reactions. [ABSTRACT FROM AUTHOR]

In the world of virtual and connected networks, the study of network's computing capabilities through graph embedding has grown in prominence. In order to implement algorithms created for the guest graph in the host graph, embedding involves simulating one architecture, called the guest, into another, called the host. In this paper, we have obtained the optimal wirelength of embedding (K 9 - C 9) n into Mesh M (3 n , 3 n) , generalized book graph GB [ 9 , 3 n - 1 , 3 n - 1 ] and extended firecracker tree EF 9 n . [ABSTRACT FROM AUTHOR]

As the urgency of PM2.5 prediction becomes increasingly ingrained in public awareness, deep-learning methods have been widely used in forecasting concentration trends of PM2.5 and other atmospheric pollutants. Traditional time-series forecasting models, like long short-term memory (LSTM) and temporal convolutional network (TCN), were found to be efficient in atmospheric pollutant estimation, but either the model accuracy was not high enough or the models encountered certain challenges due to their own structure or some specific application scenarios. This study proposed a high-accuracy, hourly PM2.5 forecasting model, poly-dimensional local-LSTM Transformer, namely PD-LL-Transformer, by deep-learning methods, based on air pollutant data and meteorological data, and aerosol optical depth (AOD) data retrieved from the Himawari-8 satellite. This research was based on the Yangtze River Delta Urban Agglomeration (YRDUA), China for 2020–2022. The PD-LL-Transformer had three parts: a poly-dimensional embedding layer, which integrated the advantages of allocating and embedding multi-variate features in a more refined manner and combined the superiority of different temporal processing methods; a local-LSTM block, which combined the advantages of LSTM and TCN; and a Transformer encoder block. Over the test set (the whole year of 2022), the model's R2 was 0.8929, mean absolute error (MAE) was 4.4523 µg/m3, and root mean squared error (RMSE) was 7.2683 µg/m3, showing great accuracy for PM2.5 prediction. The model surpassed other existing models upon the same tasks and similar datasets, with the help of which a PM2.5 forecasting tool with better performance and applicability could be established. [ABSTRACT FROM AUTHOR]

The fuzzy membership function has non-collapsability and can be used to characterize uncertainties that cannot be explored by sudden events. As the membership function does not satisfy the complementarity law, a single membership function cannot simultaneously characterize the positive and negative aspects of things. This article considers structured fuzzy data with negative uncertainty information, which is modeled using bipolar fuzzy graphs. On this basis, we consider the connectivity remainder of the model and obtain the corresponding theoretical results. In addition, we have discussed some related fractional factor issues. [ABSTRACT FROM AUTHOR]

In the present paper, it has been obtained that the fundamental group of n -dimensional Minkowski space with the time topology contains uncountably many copies of the additive group of integers and is not abelian. The result has been first proved for n = 2. Thereafter, it is extended to n > 2 by proving that loops nonhomotopic in M 2 continue to be nonhomotopic in Mn using embedding of M 2 in Mn as a retract through the projection map. [ABSTRACT FROM AUTHOR]

The technique used in studying the computational capabilities of interconnection networks and task distribution is graph embedding. Based on the recursively constructed graphs, the hypercube network is popular for its structure. Many variants of hypercube are considered in the literature. Augmented cube is considered as one of the best variants of hypercube as it holds many desirable properties like optimal routing in linear time complexity, vertex symmetricity, wide diameter and maximum connectivity. Our work deals with the exact wirelength, when augmented cube is embedded into certain tree and windmill structures. [ABSTRACT FROM AUTHOR]

The emotional aspects of learning are important but sometimes neglected elements of Universal Design for Learning (UDL). This paper discusses a multi-faceted project to embed mental wellbeing at De Montfort University, Leicester, a UK university with an established UDL policy. The project encompassed the development of teacher training to consider student variability in emotional responses to learning; the provision of tutor resources to support the embedding of learning about wellbeing within the curriculum; and approaches to sharing practice across the campus. This paper argues that academic staff are key players in the support of students' mental wellbeing and that a whole organisation approach, rather than a narrow focus on the provision of mental health support services, is an effective route to improving students' mental wellbeing. [ABSTRACT FROM AUTHOR]

We propose a unified 3D flow framework for joint learning of shape embedding and deformation for different categories. Our goal is to recover shapes from imperfect point clouds by fitting the best shape template in a shape repository after deformation. Accordingly, we learn a shape embedding for template retrieval and a flow-based network for robust deformation. We note that the deformation flow can be quite different for different shape categories. Therefore, we introduce a novel multi-hub module to learn multiple modes of deformation to incorporate such variation, providing a network which can handle a wide range of objects from different categories. The shape embedding is designed to retrieve the best-fit template as the nearest neighbor in a latent space. We replace the standard fully connected layer with a tiny structure in the embedding that significantly reduces network complexity and further improves deformation quality. Experiments show the superiority of our method to existing state-of-the-art methods via qualitative and quantitative comparisons. Finally, our method provides efficient and flexible deformation that can further be used for novel shape design. [ABSTRACT FROM AUTHOR]

ABSTRACT Class II microcins are antimicrobial peptides that have shown some potential as novel antibiotics. However, to date, only 10 class II microcins have been described, and the discovery of novel microcins has been hampered by their short length and high sequence divergence. Here, we ask if we can use numerical embeddings generated by protein large language models to detect microcins in bacterial genome assemblies and whether this method can outperform sequence-based methods such as BLAST. We find that embeddings detect known class II microcins much more reliably than does BLAST and that any two microcins tend to have a small distance in embedding space even though they typically are highly diverged at the sequence level. In data sets of Escherichia coli, Klebsiella spp., and Enterobacter spp. genomes, we further find novel putative microcins that were previously missed by sequence-based search methods.IMPORTANCEAntibiotic resistance is becoming an increasingly serious problem in modern medicine, but the development pipeline for conventional antibiotics is not promising. Therefore, alternative approaches to combat bacterial infections are urgently needed. One such approach may be to employ naturally occurring antibacterial peptides produced by bacteria to kill competing bacteria. A promising class of such peptides are class II microcins. However, only a small number of class II microcins have been discovered to date, and the discovery of further such microcins has been hampered by their high sequence divergence and short length, which can cause sequence-based search methods to fail. Here, we demonstrate that a more robust method for microcin discovery can be built on the basis of a protein large language model, and we use this method to identify several putative novel class II microcins.

Incomplete knowledge graphs are common problem that can impair task accuracy. As knowledge graphs grow extensively, the probability of incompleteness increases. Link prediction addresses this problem, but accurate and efficient link prediction methods are needed to handle incomplete and extensive knowledge graphs. This study proposed modifications to the embedding-based link prediction using clustering to improve performance. The proposed method involves four main processes: embedding, clustering, determining clusters, and scoring. Embedding converts entities and relations into vectors while clustering groups these vectors. Selected clusters are determined based on the shortest distance between the centroid and the incomplete knowledge graph. Scoring measures relation rankings, and link prediction result is selected based on highest scores. The link prediction performance is evaluated using Hits@1, Mean Rank, Mean Reciprocal Rank and prediction time on three knowledge graph datasets: WN11, WN18RR, and FB13. The link prediction methods used are TransE and ComplEx, with BIRCH as the clustering technique and Mahalanobis for short-distance measurement. The proposed method significantly improves link prediction performance, achieving accuracy up to 98% and reducing prediction time by 99%. This study provides effective and efficient solution for improving link prediction, demonstrating high accuracy and efficiency in handling incomplete and extensive knowledge graphs.

E-government merupakan sebuah konsep pemerintahan yang menyelenggarakan layanan publik secara digital yang didukung oleh teknologi informasi, sehingga menjadi lebih prima. Salah satu bentuk layanan publik digital yang umum ada di berbagai sektor adalah helpdesk. Helpdesk memungkinkan pengguna dapat bertanya atau melaporkan sesuatu untuk kemudian dijawab oleh staf. Permasalahan yang ada adalah ketersediaan staf yang hanya dapat menjawab pada jam kerja, sedangkan pengguna bisa jadi menemukan permasalahan yang urgent di luar jam kerja. Oleh karena itu, solusi yang ditawarkan adalah penerapan chatbot, sehingga dapat melayani kapan pun, sekaligus meringankan kerja dari staf. Penelitian ini mencoba untuk merancang salah satu komponen pada chatbot, yaitu model klasifikasi intensi, dengan metode Long Short-Term Memory. Data yang digunakan merupakan 501 teks riwayat pertanyaan dari database Helpdesk TIK UB yang termasuk ke dalam 7 kelas intensi yang ditentukan. Data akan melalui beberapa tahap prapemrosesan sebelum kemudian dilakukan pemodelan dan beberapa pengujian. Tahap pengujian meliputi pemilihan embedding yang digunakan, pemilihan teknik augmentasi data, dan penyetelan hyperparameter. Hasil dari keseluruhan pengujian, didapatkan model terbaik yang mampu menghasilkan akurasi sempurna untuk data latih dan data uji, serta loss 0,004 untuk data latih dan 0,044 untuk data uji. Abstract E-government is a government concept that organizes digital public services supported by information technology, so that they become more excellent. One form of digital public service that is common in various sectors is the helpdesk. Helpdesk allows users to ask questions or report something to be answered by staff. The problem that exists is the availability of staff who can only answer during working hours, while users may find urgent problems outside of working hours. Therefore, the solution offered is the implementation of a chatbot, so that it can serve at any time, while easing the work of staff. This study attempts to design one of the components of the chatbot, namely the intention classification model, using the Long Short-Term Memory method. The data used are 501 question history texts from the Helpdesk TIK UB database which fall into the 7 specified intensity classes. The data will go through several pre-processing stages before then being modeled and tested. The testing phase includes selecting the embedding used, data augmentation techniques selection, and hyperparameter tuning. The best model is obtained which can produce perfect accuracy for training data and test data, as well as a loss of 0.004 for training data and 0.044 for test data.

This paper reproduces the result of Elliot, namely that the irreducible finite dimensional representation of the Lie algebra su(3) of highest weight (m, n) is decomposed according to the embedding so(3) ⊂ su(3). First, a realisation (a representation in terms of vector fields) of the Lie algebra su(3) is constructed on a space of polynomials of three variables. The special polynomial basis of the representation space is given. In this basis, we find the highest weight vectors of the representation of the Lie subalgebra so(3) and in this way the representation space is decomposed to the direct sum of invariant subspaces. The process is illustrated by the example of the decomposition of the representation of highest weight (2, 2). As an additional result, the generating function of the decomposition is given.