Redundant Data Research Articles

A machine learning-based model for stroke prediction

In recent years, machine learning has highlighted good results in the early diagnosis and prediction of diseases. Stroke is a serious threat to human health. Early prediction of stroke is of great significance for treatment and intervention. This paper mainly investigates the application of different machine learning models in stroke prediction and compares the performance of each model. First, we collected some data, which contained 5110 entries or records and 12 different attributes. The dataset was then subjected to various preprocessing measures, such as eliminating data redundancy. Seven different machine learning models were used. Includes Logistic Regression, Support Vector Classifier (SVC, K-Neighbors Classifier, Decision Tree Classifier, Random Forest Classifier, XGBoost Classifier (XGBClassifier) and Deep Neural Networks (DNN). After model comparison, we found that when the dataset was extremely imbalanced, the AUC of DNN before feature selection was 82%, which was significantly better than other machine learning models. In addition, the characteristics of each model were analyzed to provide a reference for the selection of stroke prediction models. The results of this study provide value for the early diagnosis and intervention of stroke, and also provide new ideas for the application of machine learning algorithms in the field of stroke prediction.

A machine learning-based model for stroke prediction

In recent years, machine learning has highlighted good results in the early diagnosis and prediction of diseases. Stroke is a serious threat to human health. Early prediction of stroke is of great significance for treatment and intervention. This paper mainly investigates the application of different machine learning models in stroke prediction and compares the performance of each model. First, we collected some data, which contained 5110 entries or records and 12 different attributes. The dataset was then subjected to various preprocessing measures, such as eliminating data redundancy. Seven different machine learning models were used. Includes Logistic Regression, Support Vector Classifier (SVC, K-Neighbors Classifier, Decision Tree Classifier, Random Forest Classifier, XGBoost Classifier (XGBClassifier) and Deep Neural Networks (DNN). After model comparison, we found that when the dataset was extremely imbalanced, the AUC of DNN before feature selection was 82%, which was significantly better than other machine learning models. In addition, the characteristics of each model were analyzed to provide a reference for the selection of stroke prediction models. The results of this study provide value for the early diagnosis and intervention of stroke, and also provide new ideas for the application of machine learning algorithms in the field of stroke prediction.

An improved sensing data cleaning scheme for object localization in edge computing environment

Abstract Radio frequency identification (RFID) is widely applied due to its fast identification speed and non-contact detection. However, the identification process of RFID tags is susceptible to interference from other tags and environmental factors, resulting in inaccurate identification data. To overcome these problem, this paper proposes an improved sensing data cleaning scheme for object localization in edge computing environment. In tag level data cleaning, we use adaptive sliding window and further consider dynamic tags and read rate in continuous reading cycle to adjust the window size timely and appropriately. In the reader level data cleaning, we estimate the tag number based on Chebyshev’s inequality through Markov chain for cyclic control and optimize different time slot lengths to improve the recognition rate. We build an edge computing environment and combine the proposed tag-level cleaning method and reader-level cleaning method to form a comprehensive RFID data cleaning process. Comparative experimental results show that the RFID data cleaning method proposed in this paper can effectively reduce redundant and missing data and improve the accuracy of tag recognition.

CFNet: An infrared and visible image compression fusion network

Image fusion aims to acquire a more complete image representation within a limited physical space to more effectively support practical vision applications. Although the currently popular infrared and visible image fusion algorithms take practical applications into consideration. However, they did not fully consider the redundancy and transmission efficiency of image data. To address this limitation, this paper proposes a compression fusion network for infrared and visible images based on joint CNN and Transformer, termed CFNet. First of all, the idea of variational autoencoder image compression is introduced into the image fusion framework, achieving data compression while maintaining image fusion quality and reducing redundancy. Moreover, a joint CNN and Transformer network structure is proposed, which comprehensively considers the local information extracted by CNN and the global long-distance dependencies emphasized by Transformer. Finally, multi-channel loss based on region of interest is used to guide network training. Not only can color visible and infrared images be fused directly but more bits can be allocated to the foreground region of interest, resulting in a superior compression ratio. Extensive qualitative and quantitative analyses affirm that the proposed compression fusion algorithm achieves state-of-the-art performance. In particular, rate–distortion performance experiments demonstrate the great advantages of the proposed algorithm for data storage and transmission. The source code is available at https://github.com/Xiaoxing0503/CFNet.

XGBoost-B-GHM: An Ensemble Model with Feature Selection and GHM Loss Function Optimization for Credit Scoring

Credit evaluation has always been an important part of the financial field. The existing credit evaluation methods have difficulty in solving the problems of redundant data features and imbalanced samples. In response to the above issues, an ensemble model combining an advanced feature selection algorithm and an optimized loss function is proposed, which can be applied in the field of credit evaluation and improve the risk management ability of financial institutions. Firstly, the Boruta algorithm is embedded for feature selection, which can effectively reduce the data dimension and noise and improve the model’s capacity for generalization by automatically identifying and screening out features that are highly correlated with target variables. Then, the GHM loss function is incorporated into the XGBoost model to tackle the issue of skewed sample distribution, which is common in classification, and further improve the classification and prediction performance of the model. The comparative experiments on four large datasets demonstrate that the proposed method is superior to the existing mainstream methods and can effectively extract features and handle the problem of imbalanced samples.

Automatic damage detection and data completion for operational bridge safety using convolutional echo state networks

Structural response data might be partially missing during acquisition and transmission by a structural health monitoring (SHM) system, affecting subsequent structural damage identification. Conventional deep learning methods require a substantial amount of data and have low training efficiency. With this in mind, this paper proposes a hybrid approach for data reconstruction and damage identification using an echo state network embedded with convolutional layers. The network has a training-free reservoir layer and focuses on historical information and multi-channel spatial features. As data redundancy occurs in most measured structural vibration responses, the extended Frobenius norm-principal component analysis method is applied to select valuable samples for updating the network, which greatly enhances the training efficiency. Experimental and numerical data from a physical bridge model is used to verify the data reconstruction capability and damage identification accuracy of the proposed method, and its performance is further compared with several existing deep learning methods.

A Prototype Decision Support System for Tree Selection and Plantation with a Focus on Agroforestry and Ecosystem Services

This study presents the development and application of a prototype decision support system (DSS) for tree selection specifically for Punjab, India, a region facing challenges of low forest cover and an increasing demand for sustainable land use practices. The DSS developed using the R Shiny framework integrates ecological, social, and agro-commercial criteria to facilitate scientific knowledge decision making in tree plantation. The modules in this DSS include a tree selection tool based on comprehensive species attributes, a GIS-based tree suitability map module utilizing an Analytical Hierarchical Process (AHP), and a silviculture practice information module sourced from authoritative databases. Combining sophisticated statistical and spatial analysis, such as NMDS and AHP-GIS, this DSS mitigates data redundancy in SDM while incorporating extensive bibliographic research in dataset processing. The study highlights the necessity of fundamental niche-based suitability in comparison to realized niche suitability. It emphasizes on the importance of addressing ecosystem services, agro-commercial aspects, and enhancing silvicultural knowledge. Additionally, the study underscores the significance of local stakeholder engagement in tree selection, particularly involving farmers and other growers, to ensure community involvement and support. The DSS supports agroforestry initiatives and finds applications in urban tree management and governmental programs, emphasizing the use of scientific literature at each step, in contrast to relying solely on local knowledge.

RETRACTION: Exploiting Cu+–Na+ Ion‐exchanged and Ar/H2 Annealed Glass Matrix to Synthesize Copper Nanoparticles

AbstractRETRACTION: S. Toumi, A. Adawy, A. Quaranta and K. Farah, “Exploiting Cu+–Na+ Ion‐exchanged and Ar/H2 Annealed Glass Matrix to Synthesize Copper Nanoparticles,” Journal of the American Ceramic Society (Early View): https://doi.org/10.1111/jace.19507.The above article, published online on October 24, 2023 in Wiley Online Library (wileyonlinelibrary.com), has been retracted by agreement between the journal Editor‐in‐Chief, John C. Mauro; the American Ceramic Society; and John Wiley & Sons, Inc. The retraction has been agreed upon following an investigation into concerns raised by a third party, which revealed significant redundancy in primary data and simultaneous submission with another published article by an overlapping group of authors elsewhere. Such publishing practice is against the journal's policy and Wiley's Best Practice Guidelines on Research Integrity and Publishing Ethics. The authors were informed of the decision to retract but did not agree to the retraction or the wording.

Deep jointly optical spectral band selection and classification learning

Spectral data provide material-specific information across a broad electromagnetic wavelength range by acquiring numerous spectral bands. However, acquiring such a significant volume of data introduces challenges such as data redundancy, long acquisition times, and substantial storage capacity. To address these challenges, band selection is introduced as a strategy that focuses on only using the most significant bands to preserve spectral information for a specific task. State-of-the-art methods focus on searching for the most significant bands from previously acquired data, regardless of the optical system and the classification model. Nevertheless, some deep-learning methods, such as end-to-end frameworks, allow the design of optical systems and the learning of the classification network parameters. In this paper, we model the optical band selection as a trainable layer that is coupled with a classification network, where the parameters are learned in an end-to-end framework. To guarantee a physically implementable system, we proposed two regularization terms in the training step to promote binarization and also the number of the selected bands, as we need to provide the conditions to design the physical element where the light passes through. The proposed method provides better performance than state-of-the-art band selection methods for three different spectral datasets under the same conditions.

Data redundancy of blockchain systems in construction projects

Data redundancy of blockchain systems in construction projects

Motion-error-free calibration of event camera systems using a flashing target

Event cameras, inspired by biological vision, offer high dynamic range, excellent temporal resolution, and minimal data redundancy. Precise calibration of event camera systems is essential for applications such as 3D vision. The cessation of extra gray frame production in popular models like the dynamic vision sensor (DVS) poses significant challenges to achieving high-accuracy calibration. Traditional calibration methods, which rely on motion to trigger events, are prone to movement-related errors. This paper introduces a motion-error-free calibration method for event cameras using a flashing target produced by a standard electronic display that elicits high-fidelity events. We propose an improved events-accumulator to reconstruct gray images with distinct calibration features and develop an optimization method that adjusts camera parameters and control point positions simultaneously, enhancing the calibration accuracy of event camera systems. Experimental results demonstrated higher accuracy compared to the traditional motion-based calibration method (reprojection error: 0.03 vs. 0.96 pixels). The 3D reconstruction error remained around 0.15 mm, significantly improving over the motion-based method’s 8.00 mm. Additionally, the method’s adaptability for hybrid calibration in event-based stereovision systems was verified (e.g., with frame cameras or projectors).

Optimization of Large-Scale Sparse Matrix-Vector Multiplication on Multi-GPU Systems

Sparse matrix-vector multiplication (SpMV) is one of the important kernels of many iterative algorithms for solving sparse linear systems. The limited storage and computational resources of individual GPUs restrict both the scale and speed of SpMV computing in problem-solving. As real-world engineering problems continue to increase in complexity, the imperative for collaborative execution of iterative solving algorithms across multiple GPUs is increasingly apparent. Although the multi-GPU-based SpMV takes less kernel execution time, it also introduces additional data transmission overhead, which diminishes the performance gains derived from parallelization across multi-GPUs. Based on the non-zero elements distribution characteristics of sparse matrices and the trade-off between redundant computations and data transfer overhead, this paper introduces a series of SpMV optimization techniques tailored for multi-GPU environments and effectively enhances the execution efficiency of iterative algorithms on multiple GPUs. Firstly, we propose a two-level non-zero elements-based matrix partitioning method to increase the overlap of kernel execution and data transmission. Then, considering the irregular non-zero elements distribution in sparse matrices, a long-row-aware matrix partitioning method is proposed to hide more data transmissions. Finally, an optimization using redundant and inexpensive short-row execution to exchange costly data transmission is proposed. Our experimental evaluation demonstrates that, compared with the SpMV on a single GPU, the proposed method achieves an average speedup of 2.00x and 1.85x on platforms equipped with two RTX 3090 and two Tesla V100-SXM2, respectively. The average speedup of 2.65x is achieved on a platform equipped with four Tesla V100-SXM2.

OPTIONS attack detection in WSN using Optimized Multitask Multi-Attention Residual Shrinkage Convolutional Neural Network

The rapid increase in Internet users has made web applications essential for daily services, rendering them targets for various cyber-attacks like path traversal, zero-day attacks, and injection attacks. While traditional security measures can prevent many familiar attacks, they are often ineffective against OPTIONS attacks, which involve injecting malicious code via hyperlinks to obstruct user access to legitimate webpage content. To address this novel challenge, we propose the OAD-WSN-MMRCNN technique, leveraging an Optimized Multitask Multi-Attention Residual Shrinkage Convolutional Neural Network for OPTIONS attack detection in Wireless Sensor Networks (WSNs). This approach begins by selecting a CPU parameters dataset for attack detection, followed by pre-processing with a Variational Bayesian-Based Maximum Correntropy Cubature Kalman Filter to remove redundant data. Key features such as handles, threads, processor, context switch, deferred procedure call (DPC), interrupt delta, CPU socket, and core are extracted using a variable velocity strategy particle swarm optimization algorithm. The MMRCNN, optimized with the Tyrannosaurus Optimization Algorithm, is then employed to detect normal and OPTIONS attacks. Implemented in Python, the efficacy of OAD-WSN-MMRCNN is evaluated using metrics such as energy consumption, target window, accuracy, precision, F-measure, recall, and CPU utilization. Experimental results demonstrate that OAD-WSN-MMRCNN outperforms existing techniques, achieving a 20 % improvement in detection accuracy and a 25 % reduction in energy consumption, highlighting its effectiveness and potential for enhancing web application cyber security.

An Effective Data Fusion Methodology for Multi-modal Emotion Recognition: A Survey

Emotion recognition is a pivotal area of research with applications spanning education, healthcare, and intelligent customer service. Multimodal emotion recognition (MER) has emerged as a superior approach by integrating multiple modalities such as speech, text, and facial expressions, offering enhanced accuracy and robustness over unimodal methods. This paper reviews the evolution and current state of MER, highlighting its significance, challenges, and methodologies. We delve into various datasets, including IEMOCAP and MELD, providing a comparative analysis of their strengths and limitations. The literature review covers recent advancements in deep learning techniques, focusing on fusion strategies like early, late, and hybrid fusion. Identified gaps include issues related to data redundancy, feature extraction complexity, and real-time detection. Our proposed methodology leverages deep learning for feature extraction and a hybrid fusion approach to improve emotion detection accuracy. This research aims to guide future studies in addressing current limitations and advancing the field of MER. The main of this paper review recent methodologies in multimodal emotion recognition, analyze different data fusion techniques,identify challenges and research gaps.

Hydraulic system fault diagnosis decoupling method based on 2D time-series modeling and self-attention fusion

Hydraulic systems play a pivotal and extensive role in mechanics and energy. However, the performance of intelligent fault diagnosis models for multiple components is often hindered by the complexity, variability, strong hermeticity, intricate structures, and fault concealment in real-world conditions. This study proposes a new approach for hydraulic fault diagnosis that leverages 2D temporal modeling and attention mechanisms for decoupling compound faults and extracting features from multisample rate sensor data. Initially, to address the issue of oversampling in some high-frequency sensors within the dataset, variable frequency data sampling is employed during the data preprocessing stage to resample redundant data. Subsequently, two-dimensional convolution simultaneously captures both the instantaneous and long-term features of the sensor signals for the coupling signals of hydraulic system sensors. Lastly, to address the challenge of feature fusion with multisample rate sensor data, where direct merging of features through maximum or average pooling might dilute crucial information, a feature fusion and decoupling method based on a probabilistic sparse self-attention mechanism is designed, avoiding the issue of long-tail distribution in multisample rate sensor data. Experimental validation showed that the proposed model can effectively utilize samples to achieve accurate fault decoupling and classification for different components, achieving a diagnostic accuracy exceeding 97% and demonstrating robust performance in hydraulic system fault diagnosis under noise conditions.

Optimizing Storage Efficiency Through Pipelining and Parallelization for Enhanced Data Deduplication

Nowadays, huge devices are connected to the internet to deal with many cloud environments for storing data efficiently. Also, it is used to store big data determined in business activities. However, the data storage in the cloud avoids duplicate copies of files and reduces complexity issues. However, this platform is utilized by many users for storing the file in cloud and this creates a data redundancy problem. Also, fingerprint indexing is a major issue in data storage in the cloud. So in order to overcome this issue a novel Enhanced Arithmetic Optimization Algorithm with Mixed-Kernel-based Extreme Learning Machine-based Random Forest (EAOA-MKELM-RF) method is proposed that overcome the deduplication issues. In spite, the scalability issues are solved, and the limitations determined in fingerprint indexing. Further, the pipelining and parallelizing hash function values are utilized to address the scalability issue in the deduplication process. The data storage container organizes the programs and files into data blocks, metadata files and nonduplicate blocks. However, the proposed methods determine the similarity issues and accelerate the fingerprint queries for better scalability. The datasets employed to evaluate the proposed approach are CorrAL-100 dataset, Madelon dataset, XOR-100 dataset, Infocom’05 dataset and Linux kernel dataset. The experimentation results demonstrated that the proposed method attained a superior accuracy by 96.8% in terms of restoration accuracy, robustness and delivery ratio. Additionally, the proposed approach reduces buffer time, improving overall system performance. This validation optimized the storage resources, diminished storage costs and improved overall data management.

Imbalanced data fault diagnosis method for nuclear power plants based on convolutional variational autoencoding Wasserstein generative adversarial network and random forest

Data-driven fault diagnosis techniques are significant for the stable operation of nuclear power plants (NPPs). However, in practical applications, the fault diagnosis of NPPs usually faces imbalance data problems with small fault samples and much redundant data which results in low model training efficiency and poor generalization performance. Thus, this paper proposes a convolutional variational autoencoding gradient-penalty Wasserstein generative adversarial network with random forest (CVGR) to reduce the impact of imbalanced samples on fault diagnosis. Firstly, a feature selection method based on the random forest is used to identify the most relevant measurements and reduce the impact of redundant data on fault diagnosis. Then, variational autoencoding is introduced into gradient-penalty Wasserstein generative adversarial to effectively extract original sample features and generate high-quality samples with high rationality and diversity. In addition, the convolutional neural network is used to extract the features of mixed samples to realize intelligent fault diagnosis. Finally, several experiments based on the Fuqing Unit 2 full-scope simulator under different operating conditions are used to validate the performance of the CVGR in data enhancement and intelligent fault diagnosis. The results show that the proposed method can effectively mitigate the imbalance data problem, which gives insights into intelligent fault diagnosis of NPPs.

In-Sensor Tactile Fusion and Logic for Accurate Intention Recognition.

Touch control intention recognition is an important direction for the future development of human-machine interactions (HMIs). However, the implementation of parallel-sensing functional modules generally requires a combination of different logical blocks and control circuits, which results in regional redundancy, redundant data, and low efficiency. Here, a location-and-pressure intelligent tactile sensor (LPI tactile sensor) unprecedentedly combined with sensing, computing, and logic is proposed, enabling efficient and ultrahigh-resolution action-intention interaction. The LPI tactile sensor eliminates the need for data transfer among the functional units through the core integration design of the layered structure. It actuates in-sensor perception through feature transmission, fusion, and differentiation, thereby revolutionizing the traditional von Neumann architecture. While greatly simplifying the data dimensionality, the LPI tactile sensor achieves outstanding resolution sensing in both location (<400µm) and pressure (75Pa). Synchronous feature fusion and decoding support the high-fidelity recognition of action and combinatorial logic intentions. Benefiting from location and pressure synergy, the LPI tactile sensor demonstrates robust privacy as an encrypted password device and interaction intelligence through pressure enhancement. It can recognize continuous touch actions in real time, map real intentions to target events, and promote accurate and efficient intention-driven HMIs.

Data reconciliation of indoor positioning data: Improve position data accuracy in warehouse environment

This article focuses on improving indoor positioning data through data reconciliation. Indoor positioning systems are increasingly used for resource tracking to monitor manufacturing and warehouse processes. However, measurement errors due to noise can negatively impact system performance. Redundant measurement involves the use of multiple sensor tags that provide position data on the same resource, to identify errors in the physical environment. If we have measurement data from the entire physical environment, a map-based average measurement error can be determined by specifying the points in the examined area where measurement data should be compensated and to what extent. This compensation is achieved through data reconciliation, which improves real-time position data by considering the measurement error in the actual position as an element of the variance-covariance matrix. A case study in a warehouse environment is presented to demonstrate how discrepancies in position data from two sensor tags on forklifts can be used to identify layout-based errors. The algorithm is generally capable of handling the multi-sensor problem in the case of indoor positioning systems. The key points are as follows:•The layout-based error detection is determined with the indoor positioning system measurement error.•This article shows how redundant measurements and data reconciliation can improve the accuracy of such systems.•Improving the accuracy of position data with the layout-based error map using a data reconciliation algorithm.

Gait recognition using deep learning with handling defective data from multiple wearable sensors

Gait recognition based on multiple wearable sensors has received widespread attention in recent years. Collecting data from various types of multimodal sensors allows for a more comprehensive capture of gait information. This effectively enhances gait recognition performance. However, the increase in the number of sensors also leads to the possibility of generating defective data during the process of data collection and processing, particularly data redundancy and data anomaly. Traditional solutions often involve complex data analysis and selection, which may result in important information being filtered out and high computational demands. To address these issues, this paper proposes a new method that combines the powerful feature extraction capabilities of convolutional neural networks. To address the issue of data redundancy, the Channel Multiscale-aware (CMSA) Module is designed. It extracts information with different receptive fields from different channels of the same feature map to directly reduce similarities in information. To address the issue of data anomaly, the Feature Combination based on Local Channel Attention (FCLCA) Module is designed to select channel segments least affected by data anomaly from the feature map. Attention mechanisms are incorporated into both the CMSA and FCLCA modules to facilitate the model in adapting and learning more crucial features. Additionally, improvements are made to their pooling structures. After numerous experiments, it was demonstrated that our model achieved superior performance in gait recognition tasks.