Search

In a centrifugal pump, the clearance flow is quite common due to the existence of clearance between the casing and impeller. Apart from the clearance, the impeller speed and flow rate have a significant impact on fluid frictional torque. This study uses experimental and numerical methods to investigate these dynamics. The experimental setup includes measurements of fluid frictional torque at various levels of axial clearance (0.6 mm, 1.2 mm, and 1.8 mm), flow rates (8 L/min, 10 L/min, and 12 L/min), and impeller speeds (800 rpm, 1000 rpm, and 1200 rpm). A 3-level, 3-factor factorial design (L27) is employed to systematically examine the impact of these factors on fluid frictional torque. Response Surface Methodology (RSM) and Artificial Neural Networks (ANNs) are utilized to capture complex parameter interactions, with optimization performed using a Desirability Function (DF). The analysis reveals a significant increase in fluid frictional torque with increasing axial clearance, impeller speed, and flow rate. The optimal operational parameters for minimizing fluid frictional torque in the centrifugal pump are identified as ω = = 888.94rpm, Q = 8.026L/min and Cf = 0.619mm, achieving a minimum fluid frictional torque of 0.499 Nm [ABSTRACT FROM AUTHOR]

There is a growing trend of artificial intelligence (AI) applications in veterinary medicine, with the potential to assist veterinarians in clinical decisions. A commercially available, AI-based software program (AISP) for detecting common radiographic dental pathologies in dogs and cats was assessed for agreement with two human evaluators. Furcation bone loss, periapical lucency, resorptive lesion, retained tooth root, attachment (alveolar bone) loss and tooth fracture were assessed. The AISP does not attempt to diagnose or provide treatment recommendations, nor has it been trained to identify other types of radiographic pathology. Inter-rater reliability for detecting pathologies was measured by absolute percent agreement and Gwet's agreement coefficient. There was good to excellent inter-rater reliability among all raters, suggesting the AISP performs similarly at detecting the specified pathologies compared to human evaluators. Sensitivity and specificity for the AISP were assessed using human evaluators as the reference standard. The results revealed a trend of low sensitivity and high specificity, suggesting the AISP may produce a high rate of false negatives and may not be a good tool for initial screening. However, the low rate of false positives produced by the AISP suggests it may be beneficial as a "second set of eyes" because if it detects the specific pathology, there is a high likelihood that the pathology is present. With an understanding of the AISP, as an aid and not a substitute for veterinarians, the technology may increase dental radiography utilization and diagnostic potential. [ABSTRACT FROM AUTHOR]

The assessment of compressive strength in high-performance concrete (HPC) holds significant importance, both in practical applications and in the context of its challenging mix proportions design, which often demands numerous experimental trials to achieve the desired property such as compressive strength. This process is typically resource-intensive and time-consuming. However, notwithstanding the complexities of its design, artificial intelligence (AI) techniques have demonstrated efficacy in accurately predicting desired concrete properties by optimizing mixture proportions. This research proposes AI-based models to predict the compressive strength of HPC. In this study, twenty-six different models with six approaches, namely linear regression, regression tress, support vector machine, ensembles of trees, gaussian process regression, and neural network models, were employed. In addition, different K-folds and optimizers were used. The models, based on 152 experimental results, utilized 10 variables, including the content of cement, slag, silica fume, water, superplasticizer, fine aggregate, and coarse aggregate, in addition to the moisture content of the fine and coarse aggregates, and concrete age as inputs to predict HPC's compressive strength. 80 % of the data was utilized for training, while the remaining 20 % for testing. Model performance was assessed using R² and RMSE, MSE, and MAE. For the testing data, empirical results from the performance evaluation revealed that cubic support vector machine model using 8-k fold and grid search optimizer with principal component analysis had the worst performance, while fine regression tree model using 2-k fold and Bayesian optimization had relatively superior performance with highest correlation and least error (R2 = 0.94, RMSE = 4.15, MSE = 17.279, and MAE = 3.133) in comparison with the other developed models. [ABSTRACT FROM AUTHOR]

In recent years, the complexity of last mile freight transportation management has increased. Shippers are facing demand fragmentation and often rely on a set of external carriers for the delivery to customers. Different carriers are characterized by different costs, service level and range of delivery services. Therefore, selecting the right carrier for each delivery is crucial for shippers' efficiency and effectiveness. However, research has largely overlooked this issue. This study represents a first attempt to address the carrier selection problem through the application of machine learning. Considering the real-world data of a shipper, a machine learning model is developed and its application is simulated to assign each delivery to the most suitable carrier based on delivery features and carriers' historical lead time performance. Results show that significant performance improvements can be achieved compared with the approach currently used by the shipper, highlighting the potential for further research in this area. [ABSTRACT FROM AUTHOR]

The benefits of Integrated Management Systems (Quality, Environmental, and Safety Management) certification are often discussed. While some authors suggest that certification brings financial stability, limited proof exists on this stance, especially when aiming to build on such stability to forecast future performance. The objective of this work was to evaluate the use of Artificial Intelligence to perform KPI financial forecasting in companies with certified Integrated Management Systems (IMS) and in companies without IMS certification, and then compare the results to verify whether or not the results are similar. Using an artificial neural network (ANN) model, financial indicators were forecasted both for companies that are IMS-certified and companies that are not. Data from 50 certified organizations and 50 non-certified organizations was used to train the model. Interpretation of the results was done using relative error analysis. Results showed substantial differences, with forecasts done on data from IMS-certified companies presenting much-improved accuracy. This study shows that the use of ANN to support the prediction of financial performance indicators is feasible while contributing to a better understanding of the benefits of Integrated Systems Management certification. [ABSTRACT FROM AUTHOR]

Certain types of lesions on the laryngeal mucosa may indicate the early stages of laryngeal squamous cell carcinoma (LSCC), which constitutes 98% of malignant laryngeal tumors. This study aims to develop artificial intelligence-based methods to classify laryngeal lesions using digital images obtained through narrowband endoscopy imaging. A total of 1,320 digital images of laryngeal tissue, both healthy and lesioned, were classified into four categories. Five machine learning models were developed, utilizing conventional deep convolutional neural networks (CNNs) and capsule networks (CapsNet): VGG16, VGG19, Inception V3, CapsNet without data augmentation, and CapsNet with data augmentation. The latter used images synthetically generated by an adversarial generative network (GAN). These algorithms were implemented using the Orange visual programming software and the Colab computational platform. The inclusion of GAN-enhanced data augmentation significantly improved the performance of the CapsNet classifier across all lesion types. The CapsNet model with GAN data augmentation achieved an average recall, accuracy, and F1 score of 94.7%, marking it as the second-best performing model. The highest performance was achieved by the CNN Inception V3 model, with 97% recall, accuracy, and F1 score, facilitated through visual programming. The combination of CapsNet with GAN-based data augmentation presents a viable alternative for the classification of medical images. The use of the Orange visual programming tool enabled high classification performance--97% in both accuracy and sensitivity--at low computational costs, without the need for advanced programming skills from the user. [ABSTRACT FROM AUTHOR]

Different project management processes have been used in software engineering to support managers in keeping project costs manageable. One of the essential processes in software engineering is to accurately and reliably estimate the required effort and cost to complete the projects. The domain of software cost estimation has witnessed a prominent surge in research activities in recent years and being an evolving process, it keeps opening new avenues, each with advantages and disadvantages, making it important to work out better options. This research aims to identify the factors that influence the software effort estimation using the constructive cost model (COCOMO), and artificial neural networks (ANN) model by introducing a novel cost estimation approach, COCOMO-ANN (CANN), utilizing a partially connected neural network (PCNN) with inputs derived from calibrated values of the COCOMO model. A publicly available dataset (COCOMONASA 2), various combinations of activation functions, and layer densities have been systematically explored, employing multiple evaluation metrics such as MAE, MRE, and MMRE. In the PCNN model, the ReLU activation function and a 1000-dense layer have demonstrated better performance. While layer density generally correlates with better outcomes, this correlation is not universally applicable for all activation functions and outcomes vary across different combinations. The use of the relationships between 26 key parameters of COCOMO in PCNN produced better results than FCNN by 0.59%, achieving an MRE of 6.55 and an MMRE of 7.04. The results indicated that the CANN model (COCOMO & ANN) presented better results than existing models. [ABSTRACT FROM AUTHOR]

The research team has developed an information system based on clinical blood cell analysis and designed and implemented highly innovative algorithms. A neural network model was created based on these feature data of the blood cell population. Artificial intelligence algorithms can label susceptible populations for digestive tract cancer with an accuracy rate of over 80 %. A multi universe optimized BP neural network model was implemented based on TCGA data of common immune antigens in clinical laboratories. The working mechanism of this model is to assign values to the parameters of the BP neural network by using the process of searching for the best fitness in multiple universes. This model can predict the five-year survival rate of patients based on immunohistochemical data. Based on these data, an AI algorithm was used to develop a clinical prognostic model with an accuracy rate of over 99 %. The research team used single-cell sequencing data to locate cell subtypes in the features of immunohistochemical data, providing a biological basis for artificial intelligence models. The research team explored the potential biological mechanisms of cancer progression and occurrence based on gastrointestinal neuroendocrine products, and these algorithms have contributed to the prediction of cancer survival and incidence,team invented a simple and efficient algorithm. [ABSTRACT FROM AUTHOR]

This paper explores the application of Transfer Entropy (TE) in deep neural networks as a tool to improve training efficiency and analyze causal information flow. TE is a measure of directed information transfer that captures nonlinear dependencies and temporal dynamics between system components. The study investigates the use of TE in optimizing learning in Convolutional Neural Networks and Graph Convolutional Neural Networks. We present case studies that demonstrate reduced training times and improved accuracy. In addition, we apply TE within the framework of the Information Bottleneck theory, providing an insight into the trade-off between compression and information preservation during the training of deep learning architectures. The results highlight TE's potential for identifying causal features, improving explainability, and addressing challenges such as oversmoothing in Graph Convolutional Neural Networks. Although computational overhead and complexity pose challenges, the findings emphasize the role of TE in creating more efficient and interpretable neural models. [ABSTRACT FROM AUTHOR]

Traumatic pneumothorax is a complex condition that is challenging to diagnose, particularly in hospitals, underdeveloped areas, and during mass casualty events. This study aimed to evaluate the potential of machine learning (ML) for diagnosing and assessing traumatic pneumothorax. We extracted 33 vital signs and blood gas parameters from the MIMIC-IV database, selecting 12 clinically significant features as inputs to four ML algorithms: extreme gradient boosting (XGBoost), artificial neural network (ANN), support vector machine (SVM), and k-nearest neighbors (KNN). Five-fold cross-validation was used to train and test the models, with external validation performed on the EICU database. Model performance was evaluated using AUROC, recall, and accuracy, with SHAP interpretability employed to understand feature importance. In total, 3871 participants from the MIMIC-IV database and 22,022 participants from the EICU database were analyzed. Hemoglobin, Oxygenation Index, and pH were found to be key indicators of severe traumatic pneumothorax. XGBoost exhibited the best performance, achieving an AUROC of 0.979 (95% CI: [0.966, 0.989]) on the MIMIC-IV dataset and 0.806 (95% CI: [0.740, 0.864]) on the EICU dataset. The results suggest that ML, particularly XGBoost, is faster and more convenient than traditional imaging methods, making it well-suited for emergency or mass casualty situations. ML algorithms show promise for initial diagnosis of traumatic pneumothorax, with XGBoost demonstrating strong interpretability and robust external validation. [ABSTRACT FROM AUTHOR]

Advances in remote sensing and machine learning are enhancing cropland classification, vital for global food and water security. We used multispectral Harmonized Landsat 8 Sentinel-2 (HLS) 30-m data in an artificial neural network (ANN) multi-layer perceptron (MLP) model to classify five crop classes (cotton, alfalfa, tree crops, grapes, and others) in California's Central Valley. The ANN MLP model, trained on 2021 data from the United States Department of Agriculture's Cropland Data Layer, was validated by classifying crops for an independent year, 2022. Across the five crop classes, the overall accuracy was 74%. Producer's and user's accuracies ranged from 65% to 87%, with cotton achieving the highest accuracies. The study highlights the potential of using deep learning with HLS time series data for accurate global crop classification. [ABSTRACT FROM AUTHOR]

In the past decade, the construction speed of high-rise office towers worldwide has exhibited explosive growth. The unique morphological characteristics of high-rise office towers result in higher shape factors and relatively larger thermal loads. The traditional workflow of "design-evaluation" in the early stages of design imposes constraints on the diversity of tower morphology, the timeliness of performance evaluation, and the efficient integration of systems. Therefore, targeting the geometric characteristics of high-rise office towers, a systematically developed and universally applicable design grammar, named "Vertex-Based Polygonal Generative Grammar (VPGG)" is proposed. Additionally, a corresponding early-stage performance driven High-rise Office Tower Generative Design Framework (HOT_GDF) is introduced. Case study results demonstrate that, with the support of Artificial Neural Network, utilizing this system can not only globally explore the diversity of tower morphologies but also efficiently uncover greater energy-saving potential in complex architectural forms compared to simpler cubic forms, with an improvement of up to 7.76% during the early stages of design. Designed from the perspective of architects, the framework achieves logical, refined, and visual real-time interaction between computers and human minds during the early stages of tower design. This enhances design efficiency and facilitates design decision-making. It systematically integrates considerations for environmental performance, such as thermal load and thermal comfort, into the design process. Furthermore, it couples various aspects of morphological design with corresponding building performance, helping users in making design decisions from a rational and quantifiable perspective. This captures greater design potential, encompassing both form and performance, for high-rise office towers during the initial design phase. [Display omitted] [ABSTRACT FROM AUTHOR]

Artificial Intelligence (AI) can enhance the precision and efficiency of diagnostics and treatments in oral and maxillofacial surgery (OMS), leveraging advanced computational technologies to mimic intelligent human behaviors. The study aimed to examine the current state of AI in the OMS literature and highlight the urgent need for further research to optimize AI integration in clinical practice and enhance patient outcomes. A scoping review of journals related to OMS focused on OMS-related applications. PubMed was searched using terms "artificial intelligence", "convolutional networks", "neural networks", "machine learning", "deep learning", and "automation". Ninety articles were analyzed and classified into the following subcategories: pathology, orthognathic surgery, facial trauma, temporomandibular joint disorders, dentoalveolar surgery, dental implants, craniofacial deformities, reconstructive surgery, aesthetic surgery, and complications. There was a significant increase in AI-related studies published after 2019, 95.6% of the total reviewed. This surge in research reflects growing interest in AI and its potential in OMS. Among the studies, the primary uses of AI in OMS were in pathology (e.g., lesion detection, lymph node metastasis detection) and orthognathic surgery (e.g., surgical planning through facial bone segmentation). The studies predominantly employed convolutional neural networks (CNNs) and artificial neural networks (ANNs) for classification tasks, potentially improving clinical outcomes. [ABSTRACT FROM AUTHOR]

Cross-project defect prediction (CPDP) plays an essential role in identifying potential defects in target projects, especially those with limited historical data, using relevant information from similar source projects. The current studies focused on three main types of software metrics for CPDP: static metrics, code-change metrics, and semantic features. However, these existing CPDP studies encounter two primary challenges: class overlap due to reduced feature dimensions and multicollinearity from integrating various software metrics. To address these challenges, we propose a CPDP model based on both reduction and hybridization techniques (RH-CPDP). The proposed model uses hybrid deep neural networks as a hybridization technique to combine the essential metrics from all metric categories, addressing the issue of class overlap to enhance prediction model efficiency. Principal component analysis (PCA) was used as a reduction method to keep the number of metrics used small, focusing on influential relationships among metrics and fault proneness and avoiding the multicollinearity problem. The experimental analysis conducted using nine open-source projects from the PROMISE dataset demonstrates that RH-CPDP surpasses current CPDP methods (TCSBoost, TPTL, DA-KTSVMO, DBN, and 3SW-MSTL) regarding area under the curve (AUC) and F1-measure. These findings highlight the effectiveness of RH-CPDP in improving the performance of CPDP techniques. • The CPDP model is proposed using reduction and hybridization techniques. • Hybrid deep neural networks are used to combine key software metrics. • Principal component analysis is used to keep the number of metrics used small. • Multiple performance indicators are used for efficiency evaluation. [ABSTRACT FROM AUTHOR]

Many deep learning-based solutions proposed to automate the analysis of camera-trap images for animal counting and behaviour detection deployed image classifiers that produce image-level labels, tackle animal counting as a classification task, and use images with one animal or attribute. They also used large image datasets which are costly, time-consuming and laborious to collect and annotate, not feasible for rare/elusive species and resource-constrained projects and did not explain the generalization of the models on images with untrained image backgrounds. This study developed animal counting and behaviour detection models for wild animals using You-OnlyLook-Once (YOLO) and small-sized datasets with 20,110 camera-trap images. The study results show that using appropriate technologies including transfer learning, data augmentation and efficient data splitting methods, it is feasible to develop high-accurate and locationinvariant object detection models using small-sized image datasets. Despite high performance, the animal counting model did not perform well on crowding/interacting animals including Guineafowl, Elephant, Lion, Zebra, Wildebeest, Baboon and Giraffe and it misclassified a significant number of Wildebeest as Buffalo and Zebra, but few Buffalo and Zebra were misclassified as Wildebeest. The behaviour detection model performed well on all behaviours except interacting. Model’s poor performance on interacting is ascribed to its messy and small training set compared to other behaviour classes. The selection of an optimal confidence threshold appropriate for a particular dataset and increasing data diversity of the training set can significantly improve recall while reducing false positives. [ABSTRACT FROM AUTHOR]

Speaker diarization is the process of identifying who is speaking at different times in audio recordings. This is important in various situations, such as recording meetings, monitoring calls in call centers, or analyzing media. In this paper, examine how well different methods for speaker diarization perform in real-life scenarios. focus on two modern techniques: I-vectors and X-vectors. I-vectors are effective for automatic speaker recognition because they create compact and efficient representations of speakers using statistical models. However, they struggle in situations involving overlapping voices or background noise. On the other hand, X-vectors overcome these limitations. They use deep neural networks to create more complex and reliable representations, making them better suited for challenging conditions. To evaluate these two approaches, used standard datasets, specifically the AMI Meeting Corpus and VoxCeleb. measured their performance using two indicators: Diarization Error Rate (DER) and Jaccard Error Rate (JER). Results show that while I-vectors are less resourceintensive and work well in ideal conditions, X-vectors perform better in real-world settings where noise and overlapping speech are present. This study provides guidance for practitioners in choosing the right approach based on their needs, considering factors such as accuracy, computational costs, and reliability. [ABSTRACT FROM AUTHOR]

Wind-structures interaction has been extensively examined in the last few decades using field measurements, full scale measurements and wind tunnel testing. These experimental approaches are considered costly and time consuming. The need for a reliable analytical approach that can be used for examining wind-effects on buildings is clear. Although Computational Fluid Dynamics (CFD) is one of the other alternative numerical options yet might not reached the level of confidence to be reliably used to finalize the structural design. On the other hand, a limited number of studies have been carried out using soft computing methods to examine wind-induced loads on structures. However, its promising results, more work is still required towards achieving the full analytical prediction of wind effects on structures. This study investigates the use of different soft-computing techniques in predicting wind pressures on tall buildings. Two deep learning methods viz deep belief network (DBN) and deep neural network (DNN), and five machine learning methods namely feedforward neural network, extreme learning machine, weighted extreme learning machine, random forest, and gradient boosting machine were evaluated, and compared in predicting the design wind pressures on tall buildings. Wind tunnel datasets, used in the current study to develop the proposed computing models, were collected from testing three tall buildings having the same full-scale horizontal dimensions of (40 m and 80 m) and different heights of (80 m, 120 m and 160 m). The buildings were tested at a scale of 1:400 in urban terrain exposure. Mean and fluctuating wind pressure coefficients on the building with the height of 120 m are herein predicted using the seven computing methods and the results were compared to the corresponding measured pressures. Overall, the examined methods performed well in the wind pressure prediction process. Furthermore, the employed DNN was found to have the best performance in predicting mean and fluctuating wind pressures with the highest correlation coefficients. Hence, the DNN was also used in predicting the mean and fluctuating wind pressures on the two other buildings with heights of 80 m and 160 m. Experimental results indicate that the employed DNN model can be effectively used in predicting wind-induced pressures on tall buildings. [ABSTRACT FROM AUTHOR]

Aircraft play a major role in meeting the fast and efficient transportation needs of modern society, thanks to their advanced features. However, gas turbine engines used in aircraft have many negative effects on human health. One of the negative effects is the exhaust gases released by these engines to nature. In this study, it is discussed to present alternative models based on heuristic methods to reduce the emission values of the synthetic fuel mixture used in the combustion chamber of gas turbine engines. For this purpose, a model based on artificial neural networks (ANN) based on the back-tracking search optimisation (BSO) algorithm is proposed by using experimentally obtained emission values found in the literature. In the proposed model, the parameters of the optimum ANN structure are first determined by the BSO algorithm. Then, by using the optimum ANN structure, the most appropriate input values were found with the BSO algorithm, and the emission values were reduced. The simulation results have shown that the proposed method will be a fast and safe alternative method for reducing emission values. [ABSTRACT FROM AUTHOR]

Copyright of International Journal of InformaticsTechnologies is the property of Institute of Informatics, Gazi 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.)

Due to the complex flow interference by side hulls, load prediction has been one of the obstacles to achieve structural health monitoring and intelligent navigation of trimarans. According to the machine learning theory, a short-term prediction method towards trimaran loads is studied by using an optimized data-driven model. In the research, the monitoring data from a trimaran model test is applied for the training and testing of long short-term memory (LSTM) neural network. The impact analysis on the factors such as input length, neuron number, artificial neural network (ANN) optimizer, and output scope, are taken. To highlight the trimaran high-frequency load fluctuation and improve the prediction accuracy, the LSTM neural network combines with different signal decomposition algorithms, such as Empirical Mode Decomposition (EMD), Ensemble Empirical Mode Decomposition (EEMD), Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN), and Variational Mode Decomposition (VMD). Through the comprehensive comparison of statistical indicators at different positions and wave environment, the VMD-LSTM model is selected by considering its highest load prediction ability among these artificial neural network models. The research will improve the load prediction accuracy in the structural health monitoring systems and offer effective technical support for intelligent unmanned trimarans. [ABSTRACT FROM AUTHOR]

Road safety is a global concern, and vehicular accidents are unfortunately becoming more common. To address this issue, a Driver Assistant System using deep learning has been developed, aimed at reducing accidents by identifying traffic signs and alerting distracted drivers. The proposed system integrates Traffic Sign Detection, Classification, and a Voice Alert Module. Using You Only Look Once Version 4 (YOLO V4) for detection and a Convolutional Neural Network (CNN) for classification, it accurately recognizes signs. This system not only increases the safety of the driver and passengers, but also contributes to the overall safety of other road users. For training the Convolutional Neural Network Model and the YOLO v4 model, images from German Traffic Sign Recognition Benchmark (GTSRB) dataset are selected. To enhance the safety aspect of the system, a voice alert generation module is integrated. When a traffic sign is detected, the system generates voice alerts or warnings based on the recognized sign's meaning. These voice alerts are designed to guide and notify the driver about important traffic regulations or warnings, such as speed limits, stop signs and pedestrian crossings, promoting safer driving for all road users. [ABSTRACT FROM AUTHOR]

Twitter is the one of the biggest social media sites, where users may share their thoughts, ideas, and opinions as well as discuss current events and live tweets. In the subject of opinion mining, creating reliable and effective algorithms for sarcasm detection on Twitter is an intriguing task. Sarcasm is the use of positive language to convey depressing emotions while speaking in opposition to one's own intentions. Sarcasm is frequently employed on social networking and micro blogging platforms, where users can offend others and find it difficult to express their true feelings. The deep learning technique utilised in the current algorithms to identify these sarcastic tweets has the limitation of not being able to predict continuous variables. A novel deep learning algorithm is proposed to identify both positive and negative terms as well as sarcasm in comments. Deep neural networks are used to classify the comments into positive and negative word categories. Customers' opinions are mined using sentiment analysis to find and extract information from the text. Sarcastically stated statements from social networking sites can be quickly categorised and recognised by using VADER (Valence Aware Dictionary and Sentiment Reasoner). [ABSTRACT FROM AUTHOR]

Advances in networking technology have made Distributed Denial of Service (DDoS) attacks a real danger to today's networks. Using logical reasoning, the network flow circumstances may be classified as an attack or a routine state to mimic DDoS detection. This research builds an Artificial Intelligence (AI) system using current improvements in Detection System (DS) and Artificial Neural Network (ANN) algorithms advances. It examines User Datagram Protocol (UDP) foods, ping foods, Transmission Control Protocol (TCP) foods, and land attacks to better understand attack behavior. The categorization model for DDoS attacks is constructed using machine learning approaches. Once trained and evaluated, the model can identify unlabeled benign or malicious network data. Experiments reveal that Decision Tree (DT), Random Forest (RF), Naïve Bayes, and ANN are more accurate in separating ordinary and attack traffic. The ANN is used to extract optimal features from Internet of Things (IoT) Intrusion Detection System (IDS) data. The DS Algorithm, a new RF optimizer, is employed for effective feature selection. Performance evaluation of the resulting model called Artificial Neural Network-Random Forest (ANN-RF), is done using the "Application DDoS Layer Dataset". RF was selected because it trains faster than DT. We have got 99.998% better accuracy than 99.930% which is the most efficient accuracy from previous work in this field. As per our results, the proposed work has detected smart accuracy and can detect it in real time while keeping the network connected at the same time. Furthermore, we do thorough empirical comparisons of various optimization techniques utilizing a variety of categorization performance metrics. The results confirmed that the proposed technique had a competitive performance across all datasets. [ABSTRACT FROM AUTHOR]

Agriculture faces many challenges of precision farming, such as the need for sustainable practices, improving yields, ensuring high yields. In resolution to these challenges, the present research provides an AI-based system that enables the use of deep learning, Global Positioning System (GPS), and Geographic Information System (GIS) technologies to create a highly intelligent smart agricultural precision farming system. Its goal is to monitoring crop health and reduce disease risk, which will lead to improved resource utilization and environmentally sustainability techniques. The proposed framework addresses the urgent need for consistency in agricultural practices, especially as global agriculture deals with pressures from climate change, resource shortages, and increasing demand for food. Traditional agricultural methods for predicting and optimizing crop yields due to increasing factors affecting crop performance Not enough generative AI, especially the use of deep learning models, supports agricultural research in many cases, allowing patterns to be identified and future results to be predicted accurately. The integration of GPS and GIS allows for more accurate mapping, real-time analysis, and effective decision-making. Weather forecasting variability, resource constraints, and demand for more food are isolated from environmental influences using deep learning models, especially Artificial Neural Networks (ANN). By using large data sets, including historical crop yield performance, soil properties, and weather conditions, the system provides highly accurate crop forecasts. Generative Adversarial Networks (GANs) and You Only Look Once (YOLO) hybrid model is playing a key role in generating crop yield and growth potential under different conditions, adjusting model accuracy over time, and this combination of ANN, GANs and YOLO optimization algorithms ensures that the system continuously enhances its predictive accuracy and overall effectiveness. The proposed generative AI framework aims to deliver these improvements in agricultural production. [ABSTRACT FROM AUTHOR]

Legal question-answering systems face significant challenges due to the complexity of legal language, the need for accurate statute interpretation, and the highly referential nature of legal documents. In practice, lawyers always refer to legal citations when making their arguments. Current models only rely on question-answer pairs during training without taking advantage of legal citations from official legal documents. This study proposes a hybrid approach to enhancing legal question-answering systems. The approach involves using retrieval models to extract contextual information from legal documents. A query mechanism retrieves relevant legal citations, which are then utilized to enhance the language model. This process enriches legal knowledge and improves the model's ability to generate and interpret legal content. Integrating the retrieved legal context with the fine-tuned language model lays the groundwork for training a question-answering system based on an Encoder-Decoder architecture. This ultimately improves its performance in delivering precise and legally sound answers. We developed a manually labeled question-answering dataset focusing on the Vietnam Civil Code to demonstrate the model's effectiveness in limited resource situations. Our approach minimizes the need for extensive training data while maintaining the model's ability to capture legal topics. Our experiments demonstrate that this hybrid architecture significantly improves performance in legal question-answering tasks compared to traditional models, particularly in scenarios where the language model lacks specific knowledge or when training datasets are limited. [ABSTRACT FROM AUTHOR]

Mobile computing faces significant challenges, including limited processing power, battery life, and memory capacity, which hinder the performance of modern applications. This research aims to optimize mobile computing by introducing an offloading system tailored to enhance the performance of devices like smart phones and tablets. The system's key feature is predicting a mobile device's next visitation location, crucial for effective offloading. Future location prediction employs a modified LSTM (Long Short-Term Memory) Recurrent Neural Network model. The system, inclusive of Mobile Communication Manager, Edge Communication Manager, and Decision Engine, dynamically makes offloading decisions based on CPU usage, execution time, energy consumption, and memory usage. In evaluating the Decision Engine algorithm, practical experiments involve a source and four edge devices, measuring task processing latency, completion time, CPU utilization, memory usage, and energy consumption. Opting for a resource-rich edge for face recognition results in a notable reduction in processing time (177ms) and lower CPU utilization (22%) compared to the source device (2049ms, 75% CPU utilization). Practical experiments affirm the Decision Engine's efficacy in optimal offloading across diverse mobile applications. [ABSTRACT FROM AUTHOR]

The escalating demand for sustainable and eco-friendly energy alternatives has sparked considerable interest in Bioethanol production through microbial processes. Conventional bioethanol production via fermentation methods are often hindered by challenges such as high costs, impurities, inconsistent yields, and extended processing durations. To mitigate these issues, the integration of modern Artificial Intelligence (AI) techniques has emerged as a transformative approach, significantly simplifying the complexities associated with microbial bioethanol production. AI-driven modelling techniques, therefore, play a pivotal role in the design, management, and optimization of bioethanol production systems. Moreover, the application of advanced AI algorithms, particularly machine learning (ML) models, enhances both the efficiency and yield of bioethanol. Therefore, a subset of machine learning, artificial neural networks (ANNs), stand out for their flexibility and remarkable accuracy in solving complex processes. ANNs have proven instrumental in modelling critical stages of the bioethanol production process, including pretreatment, fermentation, and conversion. This chapter provides a comprehensive summary of dedicated studies to improving bioethanol production by implementing advanced AI methodologies across the different stages of bioethanol production. [ABSTRACT FROM AUTHOR]

Over last decade, there is a rapid advancement in networking and computing technologies that produced large volume of sensitive data. Clearly, protecting those data from intrusions and attack is of paramount importance. Researchers have proposed many cyber security solutions and tools to protect the data. One such technique for safeguarding data is the Intrusion Detection System (IDS). This research introduces a hybrid optimization-based Feature Selection (FS) and deep learning-driven categorization namely Honey Badger Optimization-Artificial Neural Network (HBO-ANN) to identify intrusions. The Honey Badger Optimization (HBO) is an optimization technique that is utilized to choose the dataset's most important features. The Artificial Neural Network (ANN) receives reduced features dataset and classifies it as benign or attack. Additionally, a wellknown CIC-IDS 2017 dataset is employed to construct and validate the suggested system. Performance metrics for assessing the effectiveness of the suggested system are the false alarm rate, Mean Squared Error (MSE), precision, accuracy and recall. The testing and training MSEs are 0.009 and 0.00317, respectively. The model's accuracy is 97.66%. The model has a precision of 98.03% and a recall of 97.18%. There is a 1.97% false alarm rate. The outcomes have been compared with bench mark models such as Grey Wolf Optimizer-Support Vector Machine (GWO-SVM), Particle Swarm Optimization-Support Vector Machine (PSO-SVM), Fuzzy Clustering-Artificial Neural Network (FC-ANN), Bidirectional Long-Short-Term-Memory (BiDLSTM) and Feed-Forward Deep Neural Network (FFDNN). As demonstrated by the experimental results, the suggested model outperforms the benchmark algorithms. [ABSTRACT FROM AUTHOR]

Purpose: Traditional classification algorithms always have an incorrect prediction. As the misclassification rate increases, the usefulness of the learning model decreases. This paper presents the development of a wisdom framework that reduces the error rate to less than 3% without human intervention. Design/methodology/approach: The proposed WisdomModel consists of four stages: build a classifier, isolate the misclassified instances, construct an automated knowledge base for the misclassified instances and rectify incorrect prediction. This approach will identify misclassified instances by comparing them against the knowledge base. If an instance is close to a rule in the knowledge base by a certain threshold, then this instance is considered misclassified. Findings: The authors have evaluated the WisdomModel using different measures such as accuracy, recall, precision, f-measure, receiver operating characteristics (ROC) curve, area under the curve (AUC) and error rate with various data sets to prove its ability to generalize without human involvement. The results of the proposed model minimize the number of misclassified instances by at least 70% and increase the accuracy of the model minimally by 7%. Originality/value: This research focuses on defining wisdom in practical applications. Despite of the development in information system, there is still no framework or algorithm that can be used to extract wisdom from data. This research will build a general wisdom framework that can be used in any domain to reach wisdom. [ABSTRACT FROM AUTHOR]

Spindle thermal errors significantly influence the machining accuracy of machine tools, necessitating precise modeling. While deep learning methods are commonly used for this purpose, their generalization ability and performance largely depend on design of the network structure and the selection of hyperparameters. To address these challenges, this study proposes a neural network model that integrates Bayesian optimization (BO) with dilated convolution neural network (DCNN). Dilated convolutions enhance traditional CNN models by using a dilation rate, which allows the convolutional kernel to cover a larger receptive field without increasing parameter count or computational cost. To prevent local optima during hyperparameter tuning, a Bayesian algorithm based on Gaussian processes (GP) is utilized, which optimizes 9 critical hyperparameters in the DCNN. Experimental results demonstrate that the proposed model achieves over 95 % accuracy in predicting radial thermal errors for both heating and cooling states in the X and Y directions. [Display omitted] • Proposed a spindle thermal error prediction model based on Bayesian hyperparameter optimization and dilated convolutional neural networks (DCNN). • Used dilated convolution mechanisms to capture broader spatial information from thermal images without increasing parameter count, thereby preserving the resolution and spatial structure of the input thermal images. • Employed Bayesian algorithms to optimize hyperparameter selection for the dilated convolutional neural network, thereby circumventing common issues of local optima associated with traditional heuristic methods. • The model's performance in experiments significantly outperformed traditional deep learning and hyperparameter optimization methods, validating the feasibility and accuracy of this approach. [ABSTRACT FROM AUTHOR]

In recent years, cardiovascular diseases have become increasingly important as a leading cause of death globally. heart failure (HF), a chronic disease affecting some 26 million people worldwide, has become a growing pandemic. Its prevention is a national and global emergency. In India, between 1.3 and 4.6 million adults suffer from HF, and despite advances in therapy and prevention, mortality and morbidity remain high, with significant costs to the healthcare system. The purpose of this study is to conduct a comparative evaluation of ML models for predicting HF. The support vector machine (SVM) and artificial neural network (ANN) models were analyzed to determine which model offers superior accuracy. A dataset from the Kaggle platform with x records and x features was used to train the models. The results indicated that the SVM model is the best predictor of HF with an accuracy of 79%, which far exceeds the ANNs with 77%. It is concluded that the machine learning (ML) method known as SVM shows outstanding and effective performance in the task of predicting heart failure. [ABSTRACT FROM AUTHOR]

Anomaly detection on industrial multivariate time-series data is an important research topic for industrial control systems. Due to the high dimensionality of industrial multivarlate time-series and the lack of labeled anomaly· samples, deep neural networks with the ability of learning temporal patterns in an unsupervised way have become the mainstream techniques, but there is still remaining limitations. First, they have not explicitly modeled the complex correlations between different dimensions. Second, they cannot make a balance between pattern deviation anomalies and single metric anomalies. Aiming at these limitations, this paper proposes AD-FGP, a framework for industrial multivariate timeseries anomaly detection. At)-FGP has two novel features. First, it explicitly learns the correlations between different dimensions using a graph neural network. Second, it fuses a generative model and a predictive model to detect both pattern deviation anomalies and single metric anomalies effectively. We conducted extensive experiments based on both real-world and public datasets. Experiment results show that AD-FGP has a best overall anomaly detection performance by increasing the F 1 -score 5% to 40% as compared to the baseline methods. [ABSTRACT FROM AUTHOR]

Globally, liver cancer ranks as the sixth most frequent malignancy cancer. The importance of early detection is undeniable, as liver cancer is the fifth most common disease in men and the ninth most common cancer in women. Recent advances in imaging, biomarker discovery, and genetic profiling have greatly enhanced the ability to diagnose liver cancer. Early identification is vital since liver cancer is often asymptomatic, making diagnosis difficult. Imaging techniques such as Magnetic Resonance Imaging (MRI), Computed Tomography (CT), and ultrasonography can be used to identify liver cancer once a sample of liver tissue is taken. In recent research, reliable detection of liver cancer with minimal computing computational complexity and time has remained a serious difficulty. This paper employs the DenseNet model to enhance the detection of liver nodules with tumors by segmenting them using UNet and VGG using Fastai (UVF) in CT images. Its dense interconnections distinguish the DenseNet between layers. These dense connections facilitate the propagation of gradients and the flow of information throughout the network, thereby enhancing the efficacy and performance of training. DenseNet's architecture combines dense blocks, bottleneck layers, and transition layers, allowing it to achieve a compromise between expressiveness and computing efficiency. Finally, the 3D liver nodular models were created using a ray-casting volume rendering approach. Compared to other state-of-the-art deep neural networks, it is suitable for clinical applications to assist doctors in diagnosing liver cancer. The proposed approach was tested on a 3Dircadb dataset. According to experiments, UVF segmentation on the 3Dircadb dataset is 97.9% accurate. According to the study, the DenseNet and UVF segment liver cancer better than prior methods. The system proposes automated 3D liver cancer tumor visualization. [ABSTRACT FROM AUTHOR]

Intrusion detection (ID) is a cyber security practice that encompasses the process of monitoring network activities to identify unauthorized or malicious actions. This includes problems like the difficulties of existing intrusion detection models to identify emerging attacks, generating many false alarms, and their inability and difficulty to adapt themselves with time when it comes to threats, hence to overcome all those existing challenges in this research develop a Prairie Araneida optimization based fused Convolutional Neural Network model (PAO-CNN) for intrusion detection. The fused CNN (Convolutional Neural Netowrk) is a remarkable development since it combines statistical features that are extracted from the processed data and provide enhanced capabilities for the model to capture complicated patterns existing in intrusion datasets. The adoption of a fused architecture represents an integrated way towards intrusion detection where the model can significantly interpret various features to achieve higher accuracy. On top of this, the Prairie Araneida stage which is based on coyote behavior and social spider colonies respectively plays a role in enabling to handling of intricate optimization landscapes. The dual contribution of a fused CNN and novel optimization strategies strengthens the research's goal to design an effective intrusion detection system that can evolve with new cyber threats. When the Training percentage (TP) is set to 90, the model's performance can be assessed using metrics like Accuracy, Sensitivity, and Specificity. In this particular dataset, these metrics reach approximately 97.78%, 96.25%, and 96.15%, respectively, which are crucial values. Additionally, when using a k-fold value of 6, the model achieves metrics of 97.04% Accuracy, 97.37% Sensitivity, and 96.48% Specificity. [ABSTRACT FROM AUTHOR]

This study presents an energy-efficient Internet of Things (IoT)-based wireless sensor network (WSN) framework for autonomous data validation in remote environmental monitoring. We address two critical challenges in WSNs: ensuring data reliability and optimizing energy consumption. Our novel approach integrates an artificial neural network (ANN)-based multi-fault detection algorithm with an energy-efficient IoT-WSN architecture. The proposed ANN model is designed to simultaneously detect multiple fault types, including spike faults, stuck-at faults, outliers, and out-of-range faults. We collected sensor data at 5-minute intervals over three months, using temperature and humidity sensors. The ANN was trained on 70% of the 26,280 data points per sensor, with 15% each for validation and testing. Our framework demonstrated a 97.1% improvement in fault detection accuracy (measured by F1 score) compared to existing methods, including rule-based, moving average, and statistical outlier detection approaches. The energy efficiency of the system was evaluated through 24-h power consumption tests, showing significant savings over traditional WSN architectures. Key contributions include a multi-fault detection ANN model balancing accuracy and computational efficiency, an energy-optimized IoT-WSN architecture for remote deployments, and a comprehensive performance evaluation framework. While our approach offers improvements in both data validation and energy efficiency, we acknowledge limitations such as potential scalability issues and the need for further real-world testing. This research advances the field of remote environmental monitoring by providing a robust, energy-efficient solution for ensuring data reliability in challenging deployment scenarios. Future work will explore more advanced machine learning techniques and extended field testing to further validate and improve the system's performance. [ABSTRACT FROM AUTHOR]

Scanning jet electrodeposition (SJE) is a promising method for electrodeposition. Integrated with automation and machine learning, intelligent electrodeposition can be achieved. In this study, a machine learning method was used to develop the relationship between the process parameters and properties of Ni-Co alloys produced by SJE. To evaluate the various models, the coefficient of determination (R2), mean absolute percentage error (MAPE), and root mean square error (RMSE) were calculated as performance metrics. The multivariate nonlinear regression (MNR) model shows a higher R2, lower MAPE and RMSE values, and better performance in both the training and testing sets, indicating superior approximation and prediction accuracies for microhardness. A back-propagation artificial neural network (BP-ANN) shows stable R2 value of closest to 1, and the lowest MAPE and RMSE for corrosion current density and presents an exceptional generalisation ability. Finally, under the present experimental conditions, the highest microhardness and corresponding process parameters were determined using the MNR model, and the lowest corrosion current density was optimised using the BP-ANN model. [ABSTRACT FROM AUTHOR]

Siamese networks utilize deep learning models to achieve a balance between tracking speed and accuracy in visual object tracking. However, in low light and other challenging lighting conditions, the contours and textures of the tracked object may not be accurately represented in the feature maps. This can result in blurriness and interference from similar distractions, as non-edge features may be overlooked. To tackle these challenges, this paper proposes a multi-feature object tracking approach that combines Siamese networks with Local Binary Patterns (LBP) and attention mechanisms. After preprocessing the video frames for LBP feature extraction, the resulting local binary pattern maps are input into a SiamRPN network, with ResNeSt as its backbone. The incorporation of local binary patterns enhances feature representation by capturing texture information from the target's appearance. Additionally, a coordinated attention mechanism is applied after each feature layer of the network to further improve tracking accuracy. This mechanism dynamically adjusts the weights of different features to optimize performance across various scenarios. The target's position is then determined by merging the deep and shallow features retrieved by the network, which are subsequently fed into the RPN network for regression classification. Experimental results demonstrate the effectiveness of the proposed algorithm in tracking objects. [ABSTRACT FROM AUTHOR]

This study explores timber defect classification using features extracted from Colour Uniform Local Binary Patterns (CULBP) for four timber species: Rubberwood, KSK, Merbau, and Meranti, evaluating eight common defects such as bark pocket, blue stain, borer holes, brown stain, knot, rot, split, and wane. A series of colour feature sets were extracted from individual and combined RGB channels, with classification performed via an Artificial Neural Network (ANN). Results demonstrate that RGB channels without extensive colour normalisation deliver optimal accuracy, achieving 90.2% - higher than all other channels. Compared to previous studies using grayscale images and other LBP variations, our approach outperformed prior models like Daubechies Wavelet and LBP (85%) and basic LBP methods (65.4%), confirming that integrating RGB colour channels significantly enhances classification. Among timber species, Rubberwood achieved the highest accuracy, with particularly strong results for Rubberwood Rot (92.2%) and Merbau Hole (95.6%), underscoring the advantages of colour-based feature extraction in timber defect classifcation. [ABSTRACT FROM AUTHOR]

Distinguishing between different diseases and insect pests that affect maize crops is difficult. Therefore, in this study, a maize disease and pest database was established to train the AlexNet, VGG16, and ResNet50 deep convolutional neural network models. After training, these models were used to identify different types of diseases via the identification of disease indicators in local images. The test results showed that the validation set recognition accuracies of the AlexNet, VGG16, and ResNet50 models were 86.98%, 87.70%, and 85.98%, respectively. Compared with the traditional method, the recognition accuracy of the proposed method was superior. This result provides a strong basis for pest control work via a real-time and accurate technique for identifying agricultural pests. [ABSTRACT FROM AUTHOR]

Effective detection of DGA (Domain Generation Algorithm) domain names is crucial for identifying and countering Botnets, and safeguarding cyber security. In this paper, we propose a new detection method using a hybrid deep neural network with multi-dimensional features. Firstly, multi-dimensional features are employed to bolster extracting the implicit semantic content inherent in DGA domain names. Secondly, a hybrid deep neural network, which integrates both CNN (Convolutional Neural Network) and BiLSTM (Bi-directional Long Short-Term Memory network), is utilized to effectively extract and synthesize the distinctive features of DGA domain names. Finally, comparison experiments are designed to evaluate the model's overall performance and detection accuracy. Experimental results demonstrate the efficacy of the proposed model. In the two-classification, we attained a precision rate of 97.72% and an impressive F1 score of 98.20%, indicative of a fine balance between precision and recall. In the multi-classification, our model still performed well, with a precision rate of 96.90% and an F1 score of 96.92%, further underscoring its robustness and adaptability. Compared to other models, our model achieved a detection rate of 100% for more DGA families. The model demonstrated powerful abilities, especially in distinguishing among different semantic features, and it exhibited particularly exceptional detection performance for DGA domain names generated with fixed lengths or fixed letter patterns. [ABSTRACT FROM AUTHOR]

Cloud computing has revolutionized the way businesses and individuals access and utilize computing resources. Efficient virtual machine placement is a critical aspect of optimizing resource utilization, reducing operational costs, energy consumption, service level agreement and minimum virtual machine migrations, execution time, and ensuring the overall performance of cloud services. This manuscript introduces a novel approach that combines the creative problem-solving capabilities of brainstorming with the computational power of Artificial Neural Networks (ANN) to address the virtual machine placement problem in cloud environments. In this study, we propose a hybrid technique that leverages the collective intelligence of human brainstorming to generate a diverse set of placement strategies. These strategies are then evaluated, refined, and optimized using an ANN model trained on historical cloud resource allocation workload logs. By integrating the human creative process with the data-driven predictive capabilities of ANN, our approach aims to overcome the limitations of traditional virtual machine placement algorithms, which often struggle to adapt to dynamic workloads and changing resource requirements. The manuscript provides a detailed description of the hybrid technique, including the process of brainstorming for generating placement strategies, data collection and preprocessing, ANN model development, and the integration of these components into an efficient placement system. We present experimental results demonstrating the effectiveness of our approach in optimizing resource allocation, improving service performance, and optimizing resource utilization, reducing energy consumption, service level agreement and minimum virtual machine migrations, reducing execution time compared to existing static, and meta-heuristic methods. The proposed Brain Storming with ANN based Hybrid Technique offers a promising solution for enhancing the efficiency of virtual machine placement in cloud computing environments. BSO-ANN outperforms the existing techniques using performance metrics (energy consumption(Kwh), execution time(ms), SLA violations, and number of migrations). It combines human ingenuity with data-driven insights to adapt to the ever-changing dynamics of cloud workloads. This manuscript contributes to the ongoing research in cloud resource management, offering a practical approach for cloud service providers and organizations to better utilize their resources and enhance the overall quality of cloud-based services. © 2012 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Global Science and Technology Forum Pte Ltd. [ABSTRACT FROM AUTHOR]

This study investigates the utilization of artificial intelligence (AI) for feature engineering in the education sector, highlighting its potential to enhance individualized learning and improve academic outcomes. The correlation analysis, performed using a correlation matrix of the feature set, indicated that specific pairings of characteristics exhibit a strong association, resulting in the ineffectiveness of conventional models. In order to tackle this issue, we utilized three sophisticated machine learning methodologies: Adaptive Lasso (ALasso), Artificial Neural Networks (ANN), and Support Vector Regression (SVR). The ALasso model discovered several influential characteristics, namely Gender (X5), Education (X1), Hours of Work (X4), and Marital Status (X6), that significantly affect salaries. Subsequently, a comparative evaluation of these methods was conducted using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). The results demonstrated that SVR outperformed the other techniques, with the most optimal RMSE of 0.595 and MAE of 0.423. These findings emphasize the significance of using data-driven strategies in policymaking and propose further investigation into the use of AI methods in various educational contexts to improve the identification of features and the performance of models. [ABSTRACT FROM AUTHOR]

Background: The evaluation of gait function and severity classification of stroke patients are important to determine the rehabilitation goal and the level of exercise. Physicians often qualitatively evaluate patients' walking ability through visual gait analysis using naked eye, video images, or standardized assessment tools. Gait evaluation through observation relies on the doctor's empirical judgment, potentially introducing subjective opinions. Therefore, conducting research to establish a basis for more objective judgment is crucial. Objective: To verify a deep learning model that classifies gait image data of stroke patients according to Functional Ambulation Category (FAC) scale. Methods: Gait vision data from 203 stroke patients and 182 healthy individuals recruited from six medical institutions were collected to train a deep learning model for classifying gait severity in stroke patients. The recorded videos were processed using OpenPose. The dataset was randomly split into 80% for training and 20% for testing. Results: The deep learning model attained a training accuracy of 0.981 and test accuracy of 0.903. Area Under the Curve(AUC) values of 0.93, 0.95, and 0.96 for discriminating among the mild, moderate, and severe stroke groups, respectively. Conclusion: This confirms the potential of utilizing human posture estimation based on vision data not only to develop gait parameter models but also to develop models to classify severity according to the FAC criteria used by physicians. To develop an AI-based severity classification model, a large amount and variety of data is necessary and data collected in non-standardized real environments, not in laboratories, can also be used meaningfully. [ABSTRACT FROM AUTHOR]

Given the high cost and complexity associated with acquiring a large number of experimental data of cutting forces, coupled with the challenges of overfitting and weak generalization in machine learning models for cutting forces prediction under small sample conditions, we propose two methods that employ the domain adaptation extreme learning machine (DAELM) algorithms to establish precise prediction models of cutting forces in small sample scenarios. In these methods, the large sample theoretical dataset of cutting forces calculated by parallel-sided shear zone model is used as the source domain dataset, while the small sample experimental dataset of cutting forces obtained by metal cutting experiments serves as the target domain dataset, and the cutting forces prediction models based on transfer learning are established employing DAELM algorithms. Applying these methods, precise prediction models of cutting forces in orthogonal cutting of 6061-T6 aluminum alloy have been established. Compared to the cutting force prediction models established using traditional neural network algorithms, those established using the proposed methods exhibit higher prediction precision and stronger generalization ability, even when only a small sample experimental dataset of cutting forces is available. The research findings can be applied to the transfer learning-based precise modeling of other continuously varying physical quantities in metal cutting processes under small sample conditions. [Display omitted] • Apply domain adaptation extreme learning machine algorithms to regression task. • Propose two cutting force modeling methods using domain adaptation extreme learning machines. • Develop precise cutting force prediction models for orthogonal cutting of 6061-T6 aluminum alloy. • Achieve high prediction precision with fewer data than traditional neural network models. • Maintain cutting force prediction error below 5 % with only 8 experimental data. [ABSTRACT FROM AUTHOR]

Background: Athletic performance can be measured with a variety of clinical and functional assessment techniques. There is a need to better understand the relationship between the brain's electrical activity and the body's physiological performance capabilities in real-time while performing physical tasks related to sport. Orthopedic functional assessments used to monitor the neuroplastic properties of the central nervous system lack objectivity and/or pertinent functionality specific to sport. The ability to assess brain wave activity with physiological metrics during functional exercises associated with sport has proven to be difficult and impractical in real-time sport settings. Quantitative electroencephalography or qEEG brain mapping is a unique, real-time comprehensive assessment of brain electrical activity performed in combination with physiometrics which offers insight to neurophysiological brain-to-body function. Brain neuroplasticity has been associated with differences in musculoskeletal performance among athletes, however comparative real-time normal data to benchmark performance capabilities is limited. Purpose/Design: This prospective, descriptive case series evaluated performance of task-driven activities using an innovative neurophysiological assessment technique of qEEG monitored neurophysiological responses to establish a comparative benchmark of performance capabilities in healthy, uninjured Division-I athletes. Methods: Twenty-eight healthy uninjured females (n=11) and males (n=17) NCAA Division-I athletes participated in real-time neurophysiological assessment using a Bluetooth, wireless 21-channel dry EEG headset while performing functional activities. Results: Uninjured athletes experienced standard and regulated fluctuations of brain wave activity in key performance indicators of attention, workload capacity and sensorimotor rhythm (SMR) asymmetries. Conclusion: qEEG neurophysiological real-time assessment concurrent with functional activities in uninjured, Division-I athletes may provide a performance capability benchmark. Real-time neurophysiological data can be used to monitor athletes' preparedness to participate in sport, rehabilitation progressions, assist in development of injury prevention programs, and return to play decisions. While this paper focuses on healthy, uninjured participants, results underscore the need to discen pre-injury benchmarks. Level of Evidence: 4 [ABSTRACT FROM AUTHOR]

Identification of nonlinear dynamic systems is a critical task in various fields. Artificial neural networks have been widely used for this purpose due to their ability to approximate complex functions. However, their computational efficiency and stability often pose challenges, especially in real-time applications. Quantum computation has shown potential for enhancing computational performance, but its integration with neural networks is still under investigation. The primary motivation addressed in this paper is the development of an effective strategy for synthesizing and applying recurrent quantum neural networks based on Lyapunov stability criteria (RQNN-LS) for nonlinear system identification. This model enhances the computational efficiency of recurrent neural networks by incorporating quantum computation into the neural network characteristics by using qubit neurons for data processing. Additionally, adaptive learning rates are derived based on Lyapunov stability theory for online tuning of the parameters to guarantee the stability of the proposed technique. The applicability and superiority of the presented RQNN-LS identifier are verified through the simulation and practical results of nonlinear system identification, comparing its performance with other existing identification techniques. The comparative results demonstrated significant improvements in computational efficiency with the proposed technique and highlighted the merits and superiority of the developed model over other methodologies. [ABSTRACT FROM AUTHOR]

In recent decades, automatic license plate recognition (ALPR) has emerged as a critical application of artificial intelligence in intelligent transportation systems (ITS), addressing a multitude of existing issues. Several algorithms have been developed to improve ALPR accuracy under a variety of conditions, each with its own set of strengths and limitations. The KSA intends to digitize traditional services as part of its Vision 2030 initiative. In this methodology, the ALPR system for accurately recognizing KSA's license plates (LP) leverages deep neural networks (DNNs) as a powerful technique, capable of performing effectively in unconstrained environments with images taken under various conditions, even when the training data is limited. The approach employs YOLOv8 to detect LPs and alphanumeric characters in real-time, followed by using a convolutional neural network (CNN) for character recognition. According to our finding, the YOLOv8 model outperforms other models in the identification of LPs, achieving remarkable accuracy and F1 scores (mAP@0.5 = 0.96 and mAP@0.95 = 0.97). As a result, YOLOv8 was chosen for SLP character recognition. This strategy is appropriate for ITS since it tackles a number of issues, like decreasing car theft and improving public safety. Furthermore, a locally interpretable agnostic model (LIME) offers justifications that help users better understand the model. The system's provision of clear explanations for its decisions not only satisfies performance standards but also corresponds with the increasing need for explainable AI, guaranteeing an efficient and responsible implementation. [ABSTRACT FROM AUTHOR]

Learning molecular representation is a crucial task in the field of drug discovery, particularly for various specific applications such as predicting molecular properties. Current methods are mainly based on deep neural network models, such as convolutional neural networks (CNN), recurrent neural networks (RNN), graph neural networks (GNN) and their mixed models. However, these neural network models mentioned above do not provide detailed explanations for the ability to learn molecular representations and why such experimental results occurred. In this paper, we aim to compare the performance of these models in predicting molecular properties based on molecular representation, and give more insight into deep neural network architectures for specific molecular tasks. Our experimental results demonstrate that the graph neural network can obtain superior performance on the regression tasks, while the mixed deep neural network models show better performance on the classification tasks. Ablation study also gives more explanation and analysis to the experimental results. [ABSTRACT FROM AUTHOR]

Despite the effectiveness of deep neural networks, feed-forward neural networks (FFNNs) continue to play a crucial role in many applications, especially when dealing with limited data availability. The primary challenge in FFNNs is determining the optimal weights during the training process, aiming to minimise the disparity between actual and predicted outputs. Although gradient-based techniques like backpropagation (BP) have traditionally been popular for FFNN training, they come with inherent limitations, such as sensitivity to initial weights and susceptibility to getting trapped in local optima. To overcome these challenges, we introduce a novel approach based on the Gaining-Sharing Knowledge-based(GSK) algorithm. To the best of our knowledge, this paper represents the first exploration of GSK for neural network training. After obtaining the appropriate weights for the FFNN by the GSK, the weights and biases are utilised to initialise a Levenberg–Marquardt backpropagation (LMBP) algorithm, serving as a local search component. In other words, our proposed algorithm, GSK-LocS, leverages the global search capabilities of the GSK algorithm and combines them with the local search capabilities of LMBP for neural network training. This integration mitigates sensitivity to initial values and reduces the risk of being trapped in local optima. Experimental results conducted on classification and approximation problems provide compelling evidence that our proposed algorithm is highly competitive compared to other existing methods. [ABSTRACT FROM AUTHOR]