Performance Improvement Research Articles

Discrete Morse Sandwich: Fast Computation of Persistence Diagrams for Scalar Data - An Algorithm and A Benchmark.

This paper introduces an efficient algorithm for persistence diagram computation, given an input piecewise linear scalar field f defined on a d-dimensional simplicial complex K, with d ≤ 3. Our work revisits the seminal algorithm "PairSimplices" [31], [103] with discrete Morse theory (DMT) [34], [80], which greatly reduces the number of input simplices to consider. Further, we also extend to DMT and accelerate the stratification strategy described in "PairSimplices" [31], [103] for the fast computation of the 0th and (d-1)th diagrams, noted D0(f) and Dd-1(f). Minima-saddle persistence pairs ( D0(f)) and saddle-maximum persistence pairs ( Dd-1(f)) are efficiently computed by processing , with a Union-Find , the unstable sets of 1-saddles and the stable sets of (d-1)-saddles. We provide a detailed description of the (optional) handling of the boundary component of K when processing (d-1)-saddles. This fast pre-computation for the dimensions 0 and (d-1) enables an aggressive specialization of [4] to the 3D case, which results in a drastic reduction of the number of input simplices for the computation of D1(f), the intermediate layer of the sandwich. Finally, we document several performance improvements via shared-memory parallelism. We provide an open-source implementation of our algorithm for reproducibility purposes. We also contribute a reproducible benchmark package, which exploits three-dimensional data from a public repository and compares our algorithm to a variety of publicly available implementations. Extensive experiments indicate that our algorithm improves by two orders of magnitude the time performance of the seminal "PairSimplices" algorithm it extends. Moreover, it also improves memory footprint and time performance over a selection of 14 competing approaches, with a substantial gain over the fastest available approaches, while producing a strictly identical output. We illustrate the utility of our contributions with an application to the fast and robust extraction of persistent 1-dimensional generators on surfaces, volume data and high-dimensional point clouds.

Attention Mechanisms in Clinical Text Classification: A Comparative Evaluation.

Attention mechanisms are now a mainstay architecture in neural networks and improve the performance of biomedical text classification tasks. In particular, models that perform automated medical encoding of clinical documents make extensive use of the label-wise attention mechanism. A label-wise attention mechanism increases a model's discriminatory ability by using label-specific reference information. This information can either be implicitly learned during training or explicitly provided through embedded textual code descriptions or information on the code hierarchy; however, contemporary studies arbitrarily select the type of label-specific reference information. To address this shortcoming, we evaluated label-wise attention initialized with either implicit or explicit label-specific reference information against two common baseline methods-target-attention and text-encoder architecture-specific methods-to generate document embeddings across four text-encoder architectures-a convolutional neural network, two recurrent neural networks, and a transformer. We also present an extension of label-wise attention that can embed the information on the code hierarchy. We performed our experiments on the MIMIC III dataset, which is a standard dataset in the clinical text classification domain. Our experiments showed that using pretrained reference information and the hierarchical design helped improve classification performance. These performance improvements had less impact on larger datasets and label spaces across all text-encoder architectures. In our analysis, we used an attention mechanism's energy scores to explain the perceived differences in performance and interpretability between the text-encoder architectures and types of label-attention.

Cache Update and Delivery of Dynamic Contents: A Stochastic Game Approach

In this paper, we propose a cache update policy for wireless networks considering dynamic popularity for file requests. Our network scenario is a wireless network comprised of some cache-equipped access points (APs) which are deployed densely in an area and have connections with core network servers. We model the dynamics of file requests as Markov-modulated Poisson processes. Considering the congestion-dependent delay of APs' services and their overlapping coverage areas, we model the cache service for the cache-missed requests, as a stochastic game in which a Vickrey–Clarke–Groves (VCG) mechanism is exploited at each stage of the game as our proposed file routing policy, and the APs decide independently how to update their caches regarding the routed files. To this end, the APs' utilities are defined based on long-term average file delivery delay including the queueing delay in APs. By finding the Nash equilibrium of the game, we propose policies for both delivery and cache update of dynamic contents. Finally, by comparing numerical results of our proposed scheme with some conventional caching schemes, we show significant improvements in network performance in terms of file delivery delay in different conditions.

Ambulatory Long Block: A Model of Precision Education and Assessment for Internal Medicine Residents.

High-quality precision education (PE) aims to enhance outcomes for learners and society by incorporating longitudinal data and analytics to shape personalized learning strategies. However, existing educational data collection methods often suffer from fragmentation, leading to gaps in understanding learner and program performance. In this article, the authors present a novel approach to PE at the University of Cincinnati, focusing on the Ambulatory Long Block, a year-long continuous ambulatory group-practice experience. Over the last 17 years, the Ambulatory Long Block has evolved into a sophisticated data collection and analysis system that integrates feedback from various stakeholders, as well as learner self-assessment, electronic health record utilization information, and clinical throughput metrics. The authors detail their approach to data prioritization, collection, analysis, visualization, and feedback, providing a practical example of PE in action. This model has been associated with improvements in both learner performance and patient care outcomes. The authors also highlight the potential for real-time data review through automation and emphasize the importance of collaboration in advancing PE. Generalizable principles include designing learning environments with continuity as a central feature, gathering both quantitative and qualitative performance data from interprofessional assessors, using this information to supplement traditional workplace-based assessments, and pairing it with self-assessments. The authors advocate for criterion referencing over normative comparisons, using user-friendly data visualizations, and employing tailored coaching strategies for individual learners. The Ambulatory Long Block model underscores the potential of PE to drive improvements in medical education and health care outcomes.

Few-Shot Learning for Cross-Target Stance Detection by Aggregating Multimodal Embeddings

Despite the increasing popularity of the stance detection task, existing approaches are predominantly limited to using the textual content of social media posts for the classification, overlooking the social nature of the task. The stance detection task becomes particularly challenging in cross-target classification scenarios, where even in few-shot training settings the model needs to predict the stance toward new targets for which the model has only seen a few relevant samples during training. To address the cross-target stance detection in social media by leveraging the social nature of the task, we introduce cross-target text-net (CT-TN), a novel model that aggregates multimodal embeddings derived from both textual and network features of the data. We conduct experiments in a few-shot cross-target scenario on six different combinations of source-destination target pairs. By comparing CT-TN with state-of-the-art cross-target stance detection models, we demonstrate the effectiveness of our model by achieving average performance improvements ranging from 11% to 21% across different baseline models. Experiments with different numbers of shots show that CT-TN can outperform other models after seeing 300 instances of the destination target. Further, ablation experiments demonstrate the positive contribution of each of the components of CT-TN toward the final performance. We further analyze the network interactions between social media users, which reveal the potential of using social features for cross-target stance detection.

Mutual coupling reduction between antennas array for 5G mobile applications

This paper introduces the design of a multiple-input-multiple-output (MIMO) antenna optimized for low-profile applications supporting sub-6 GHz fifth-generation (5G) wireless applications. We have started the design from a single antenna with a square patch shape, each antenna array is composed from 4-element radiators fed by using power dividers and quarter microstrip lines. Mounted on a single Rogers RT5880 substrate, the MIMO antenna functions at 3.5 GHz. In order to miniature and to decrease the mutual coupling between the both antennas array we have optimised a magnetic wall based on periodic structures permitting to decrease the mutual coupling between the both antenna array. The unit element from the wall was optimised, studied and validated in order to absorb the surface current and to enhance the isolation between the different radiating elements. The dimensions of the proposed MIMO antenna are 154×220×0.578 mm³. The MIMO antenna final circuit achieves a peak gain of 9 dBi and an isolation around -30 dB. The introduction of the magnetic wall permits to enhance the isolation between the antenna array from -20 dB to -30 dB at 3.5 GHz band. This advancement contributes to the overall performance improvement of the MIMO antenna system.

WalkGAN: Network Representation Learning With Sequence-Based Generative Adversarial Networks.

Network representation learning, also known as network embedding, aims to learn the low-dimensional representations of vertices while capturing and preserving the network structure. For real-world networks, the edges that represent some important relationships between the vertices of a network may be missed and may result in degenerated performance. The existing methods usually treat missing edges as negative samples, thereby ignoring the true connections between two vertices in a network. To capture the true network structure effectively, we propose a novel network representation learning method called WalkGAN, where random walk scheme and generative adversarial networks (GAN) are incorporated into a network embedding framework. Specifically, WalkGAN leverages GAN to generate the synthetic sequences of the vertices that sufficiently simulate random walk on a network and further learn vertex representations from these vertex sequences. Thus, the unobserved links between the vertices are inferred with high probability instead of treating them as nonexistence. Experimental results on the benchmark network datasets demonstrate that WalkGAN achieves significant performance improvements for vertex classification, link prediction, and visualization tasks.

Dense Contrastive-based Federated Learning for Dense Prediction Tasks on Medical Images.

Deep learning (DL) models have achieved remarkable success in various domains. But training an accurate DL model requires large amounts of data, which can be challenging to obtain in medical settings due to privacy concerns. Recently, federated learning (FL) has emerged as a promising solution that shares local models instead of raw data. However, FL in medical settings faces challenges of client drift due to the data heterogeneity across dispersed institutions. Although there exist studies to address this challenge, they mainly focus on the classification tasks that learn global representation of an entire image. Few have been studied on the dense prediction tasks, such as object detection. In this study, we propose dense contrastive-based federated learning (DCFL) tailored for dense prediction tasks in FL settings. DCFL introduces dense contrastive learning to FL, which aligns the local optimization objectives towards the global objective by maximizing the agreement of representations between the global and local models. Moreover, to improve the performance of dense target prediction at each level, DCFL applies multi-scale contrastive representation by utilizing multi-scale representations with dense features in contrastive learning. We evaluated DCFL on a set of realistic datasets for pulmonary nodule detection. DCFL demonstrates an overall performance improvement compared with the other federated learning methods in heterogeneous settings-improving the mean average precision by 4.13% and testing recall by 6.07% in highly heterogeneous settings.

Smooth Low-Speed Sensorless Control of Switched Reluctance Machine With Heavy Load

For the switched reluctance machine (SRM) with heavy load, the inductance of the conduction phase is nonlinear, which makes the sensorless control method based on the linear region of the phase inductance no longer applicable. Meanwhile, in the case of light or no load, the estimated rotor position will abruptly change at the connection points between neighboring segments, which has an impact on the estimation of the speed. In this paper, a low-speed sensorless control method for the SRM with heavy load is proposed. The idle-phase inductance is modeled over a wide rotor position range for position estimation, and a proportional gain decreasing type second-order generalized integrator based on a frequency-locked loop (SOGI-FLL) is constructed to estimate the speed while performing rotor position filtering. The effectiveness of the proposed method is experimentally verified with a 3-phase 12/8-pole SRM prototype, and the results show that the estimated rotor position and speed are accurate and smooth, which is beneficial to performance improvement of the sensorless control.

Linguistic Models: Optimization With the Use of Conditional Fuzzy C-Means

Most fuzzy models are just numeric. In this study, we revisit, explore and augment a concept of linguistic models, viz., fuzzy models producing results that are information granules, and, specifically, intervals or fuzzy sets. The proposed architecture is formed by constructing a network of linked fuzzy sets (information granules) ininput and output spaces with the aid of a context-based Fuzzy C-Means clustering method. The user centricity of such clustering method is implied by the explicit formulation of fuzzy sets in the output space. The resulting information granules constructed in the input space are conditioned by the corresponding fuzzy sets in the output space. This arrangement can increase the interpretability of the model and represent the model as a collection of logically arranged associations among information granules. The model's overall design process is discussed along with a detailed algorithmic structure. Its experimental evaluations are provided by using both synthetic and publicly datasets. For the former, the model brings the performance improvement ranging from 91% to 250% over the models with information granules uniformly distributed in output space. For the latter, such improvement ranges from 6% to 94%. Finally, a thorough discussion is provided together with guidelines on how to develop such a linguistic model in different contexts.

Analysis and Comparison of Machine Learning Techniques for DDoS Attack Classification in Network Environments

This research presents a comparative analysis of machine learning techniques for classifying Distributed Denial of Service (DDoS) attacks within network traffic. We evaluated the performance of three algorithms: Logistic Regression, Decision Tree, and Random Forest, including their scaled-feature counterparts. The study utilized a robust methodology incorporating advanced data preprocessing, feature engineering, and Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance. The models were rigorously tested using a cross-validation framework, assessing their accuracy, precision, recall, and F1 score. Results indicated that the Random Forest algorithm outperformed the others, demonstrating superior predictive accuracy and consistency, albeit with higher computational costs. Logistic Regression, when feature-scaled, showed significant improvement in performance, highlighting the importance of data normalization in models sensitive to feature scaling. Decision Trees provided a quick and interpretable model, though slightly less accurate than the Random Forest. The research findings highlight the trade-offs between predictive performance and computational efficiency in selecting machine learning models for cybersecurity applications. The study contributes to the cybersecurity domain by elucidating the efficacy of ensemble techniques in DDoS attack classification and underscores the potential for model improvement through scaling and data balancing.

HOW TO INCREASE SUSTAINABLE RURAL TOURISM PERFORMANCE? AN EMPIRICAL STUDY IN INDONESIA

The primary objective of this study is to examine the influence of social capital, government support, and entrepreneurial orientation on the improvement of sustainable rural tourism performance. Additionally, the study aims to identify a suitable model for enhancing the overall performance of sustainable rural tourism. Using a quantitative model, this study utilized the responses of 400 rural tourism administrators. SmartPLS 3.2 is used as a tool to analyze the collected data. Smart Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to assess the interrelationships among social capital, government support, entrepreneurial orientation, and sustainable rural tourism performance. The findings show a positive impact of social capital and entrepreneurial orientation on sustainable rural tourism performance, but government support does not impact on sustainable rural tourism performance. Furthermore, social capital and government support have a positive impact on entrepreneurial orientation. This study suggests the role of entrepreneurial orientation to increase the influence of social capital and government support on sustainable rural tourism performance. Finally, This research offers a conceptual contribution to authors investigating the sustainable performance of rural tourism. Additionally, it provides practical insights for stakeholders in rural tourism, aiding them in enhancing the sustainable performance of this sector.

A dense and U-shaped transformer with dual-domain multi-loss function for sparse-view CT reconstruction.

CT image reconstruction from sparse-view projections is an important imaging configuration for low-dose CT, as it can reduce radiation dose. However, the CT images reconstructed from sparse-view projections by traditional analytic algorithms suffer from severe sparse artifacts. Therefore, it is of great value to develop advanced methods to suppress these artifacts. In this work, we aim to use a deep learning (DL)-based method to suppress sparse artifacts. Inspired by the good performance of DenseNet and Transformer architecture in computer vision tasks, we propose a Dense U-shaped Transformer (D-U-Transformer) to suppress sparse artifacts. This architecture exploits the advantages of densely connected convolutions in capturing local context and Transformer in modelling long-range dependencies, and applies channel attention to fusion features. Moreover, we design a dual-domain multi-loss function with learned weights for the optimization of the model to further improve image quality. Experimental results of our proposed D-U-Transformer yield performance improvements on the well-known Mayo Clinic LDCT dataset over several representative DL-based models in terms of artifact suppression and image feature preservation. Extensive internal ablation experiments demonstrate the effectiveness of the components in the proposed model for sparse-view computed tomography (SVCT) reconstruction. The proposed method can effectively suppress sparse artifacts and achieve high-precision SVCT reconstruction, thus promoting clinical CT scanning towards low-dose radiation and high-quality imaging. The findings of this work can be applied to denoising and artifact removal tasks in CT and other medical images.

Microstructure evolution and mechanical properties of Al–Cu–Mn–Cd alloy fabricated by CMT-wire arc additive manufacturing

The high-performance additive manufacturing of relatively small aluminum alloy components has been widely proven successful. However, wire arc additive manufacturing (WAAM) has always been challenging in achieving uniform microstructure and high mechanical properties when manufacturing large-sized parts. Therefore, this study used cold metal transition wire arc additive manufacturing (CMT-WAAM) to prepare Al-5.49Cu-0.4Mn-0.29Cd (ACMC) alloy to obtain a uniform microstructure and improve its mechanical properties. The results show that the as-deposited (AD) samples have an obvious layered deposited structure. The grains in the microstructure are all equiaxed crystals and a large amount of copper-rich eutectic phase precipitates at the grain boundaries. After heat treatment, the vast majority of the eutectic phase dissolves, and θ'phase, with a size of 100–200 nm, uniformly precipitates in the α-Al matrix. The ultimate tensile strength (UTS) and yield strength (YS) of the heat-treated (HT) specimen can reach 483.7 MPa and 412.3 MPa, respectively. There is no significant difference in the mechanical properties between the transverse and longitudinal directions of thin-walled samples. Based on experimental results and theoretical analysis, the strengthening mechanism and fracture mechanism of CMT-WAAM aluminum alloy were revealed. This work provides a theoretical basis for the microstructure optimization and mechanical performance improvement of ACMC alloy, which is of great significance for the high-performance preparation of large-sized components in the aerospace field.

Improving model performance of shortest‐path‐based centrality measures in network models through scale space

SummaryThe quality of the solution in resolving a complex network depends on either the speed or accuracy of the results. While some health studies prioritize high performance, fast algorithms are favored in scenarios requiring rapid decision‐making. A comprehensive understanding of the problem necessitates a detailed analysis of the network and its individual components. Betweenness Centrality (BC) and Closeness Centrality (CC) are commonly employed measures in network studies. This study introduces a new strategy to compute BC and CC that assesses their sensitivity in the scale space while measuring the shortest path. The scale space is generated by incorporating a scale parameter that is shown to achieve up to 60% performance improvements for various datasets. The study provides in‐depth insights into the importance of the scale space analysis. Finally, a flexible measurement tool is provided that is suitable for various types of problems. To demonstrate the flexibility and applicability, we experimented with two methods for 10 different graphs using the proposed approach.

Empowering Graph Neural Networks with Block-Based Dual Adaptive Deep Adjustment for Drug Resistance-Related NcRNA Discovery.

Drug resistance to chemotherapeutic agents remains a formidable challenge in cancer treatment, significantly impacting treatment efficacy. Extensive research has exposed the intimate involvement of noncoding RNAs (ncRNAs) in conferring resistance to cancer drugs. Understanding the intricate associations between ncRNAs and drug resistance is of pivotal importance in advancing clinical interventions and expediting drug development. However, traditional biological experimental methods are hampered by limitations, such as labor intensiveness, time consumption, and constraints in scalability. Addressing these challenges necessitates the development of efficient computational methods for the accurate prediction of potential ncRNA-drug resistance associations (NDRA). However, most existing predictive models primarily focus on known ncRNA-drug resistance associations, often neglecting the critical aspect of similarity information between ncRNAs and drug resistance. This oversight may hinder the accuracy of characterizing these associations. To overcome the limitations of existing computational models, we proposed B-NDRA, a computational framework designed for the discovery of drug resistance-related ncRNA. Initially, we constructed a heterogeneous graph that integrates ncRNA-drug resistance pairs, leveraging both known associations and similarity fusion information between ncRNAs and drug resistance. Subsequently, we employed an attention mechanism to aggregate local features of graph nodes following a dimensionality reduction of node features. Further, a graph neural network (GNN) facilitated the learning of global node embeddings. Notably, the integration of dual adaptive deep adjustment architectures, encompassing intrablock and interblock methodologies, enabled efficient extraction of global features while balancing local and global features. Finally, B-NDRA employed a multilayer perceptron to predict associations between ncRNAs and drug resistance. Through rigorous 5-fold cross-validation, B-NDRA achieved average AUC, AUPR, Accuracy, Precision, Recall, and F1-score values of 92.2%, 91.9%, 84.88%, 86.9%, 82.37%, and 84.44%, respectively. Furthermore, comparative evaluations were conducted on established models, namely, GAEMDA, GRPAMDA, and LRGCPND. The results, obtained through three distinct 5-fold cross-validation strategies, demonstrated a notable performance improvement across almost all metrics for our B-NDRA. Specific case studies targeting Doxorubicin and Imatinib further validated the practicality of our B-NDRA in discovering potential NDRA. These results confirm the potential of our B-NDRA as a valuable tool in advancing cancer research and therapeutic development. The source code and data set of B-NDRA can be found at https://github.com/XuanLi1145/B-NDRA.

DVSAI: Diverse View-Shared Anchors Based Incomplete Multi-View Clustering

In numerous real-world applications, it is quite common that sample information is partially available for some views due to machine breakdown or sensor failure, causing the problem of incomplete multi-view clustering (IMVC). While several IMVC approaches using view-shared anchors have successfully achieved pleasing performance improvement, (1) they generally construct anchors with only one dimension, which could deteriorate the multi-view diversity, bringing about serious information loss; (2) the constructed anchors are typically with a single size, which could not sufficiently characterize the distribution of the whole samples, leading to limited clustering performance. For generating view-shared anchors with multi-dimension and multi-size for IMVC, we design a novel framework called Diverse View-Shared Anchors based Incomplete multi-view clustering (DVSAI). Concretely, we associate each partial view with several potential spaces. In each space, we enable anchors to communicate among views and generate the view-shared anchors with space-specific dimension and size. Consequently, spaces with various scales make the generated view-shared anchors enjoy diverse dimensions and sizes. Subsequently, we devise an integration scheme with linear computational and memory expenditures to integrate the outputted multi-scale unified anchor graphs such that running spectral algorithm generates the spectral embedding. Afterwards, we theoretically demonstrate that DVSAI owns linear time and space costs, thus well-suited for tackling large-size datasets. Finally, comprehensive experiments confirm the effectiveness and advantages of DVSAI.

DTMFormer: Dynamic Token Merging for Boosting Transformer-Based Medical Image Segmentation

Despite the great potential in capturing long-range dependency, one rarely-explored underlying issue of transformer in medical image segmentation is attention collapse, making it often degenerate into a bypass module in CNN-Transformer hybrid architectures. This is due to the high computational complexity of vision transformers requiring extensive training data while well-annotated medical image data is relatively limited, resulting in poor convergence. In this paper, we propose a plug-n-play transformer block with dynamic token merging, named DTMFormer, to avoid building long-range dependency on redundant and duplicated tokens and thus pursue better convergence. Specifically, DTMFormer consists of an attention-guided token merging (ATM) module to adaptively cluster tokens into fewer semantic tokens based on feature and dependency similarity and a light token reconstruction module to fuse ordinary and semantic tokens. In this way, as self-attention in ATM is calculated based on fewer tokens, DTMFormer is of lower complexity and more friendly to converge. Extensive experiments on publicly-available datasets demonstrate the effectiveness of DTMFormer working as a plug-n-play module for simultaneous complexity reduction and performance improvement. We believe it will inspire future work on rethinking transformers in medical image segmentation. Code: https://github.com/iam-nacl/DTMFormer.

Promoting pain coping skills in haemophilia: A remote intervention integrating exercise and pain education.

Chronic joint pain is a significant and widespread symptom in people with haemophilia (PWH). Despite medical advancements, effective pain management remains challenging. This study presents an innovative approach that integrates remote physical exercises, pain neuroscience education, and coping strategies to address chronic pain in PWH. The remote intervention consisted of sixteen 5-min videos encompassing physical exercises for chronic pain management and pain education strategies. These videos formed an 8-week remote intervention program. Clinical and physical assessments were conducted before and after the intervention. A total of thirty-one PWHs, with a median age of 34 years (ranging from 16 to 59years), completed the remote intervention. The study revealed significant improvements in pain intensity, disability, and physical performance among PWH with chronic pain. Enhanced functional capacity was evident in the Timed Up and Go and Single Leg Stance tests, accompanied by improved scores on the Functional Independence Score in Haemophilia (FISH). Although lacking a control group, our findings are consistent with other successful exercise and pain education programs. This innovative intervention holds promise for managing chronic pain in PWH, underscoring patient empowerment, education, and collaboration. Notably, our study stands out by uniquely combining pain education and coping strategies, bolstering evidence for effective pain management.

Causal Representation Learning via Counterfactual Intervention

Existing causal representation learning methods are based on the causal graph they build. However, due to the omission of bias within the causal graph, they essentially encourage models to learn biased causal effects in latent space. In this paper, we propose a novel causally disentangling framework that aims to learn unbiased causal effects. We first introduce inductive and dataset biases into traditional causal graph for the physical concepts of interest. Then, we eliminate the negative effects from these two biases by counterfactual intervention with reweighted loss function for learning unbiased causal effects. Finally, we employ the causal effects into the VAE to endow the latent representations with causality. In particular, we highlight that removing biases in this paper is regarded as a part of learning process for unbiased causal effects, which is crucial for causal disentanglement performance improvement. Through extensive experiments on real-world and synthetic datasets, we show that our method outperforms different baselines and obtains the state-of-the-art results for achieving causal representation learning.