The significant growth and use of IoT devices, especially in smart homes, increase the risk of cyber-attacks compromising these devices and their services, resulting in data breaches and privacy invasion. Machine learning technology has been considered a sufficient security solution in IoT due to its effectiveness in detecting unsolicited traffic on an IoT network, ensuring the confidentiality, integrity, and availability of IoT devices, and, most importantly, protecting the users’ privacy. In this research, different classification-based machine learning algorithms were leveraged to detect and classify different types of network traffic on the IoT network. At the same time, the effectiveness of these classification machine learning algorithms, Decision Tree (J48), Bayes Net, and Naive Bayes were conducted on our smart home dataset. Before the implementation of this algorithm, a man-in-the-middle attack was introduced on the IoT network while the network traffic was equally captured in the process. Overall, the algorithms successfully evaluated the captured network traffic, detected the introduced MITM attack on IoT devices, and classified the traffic into solicited and unsolicited traffic. According to the findings, the Naive Bayes algorithm outperformed the others with an accuracy of 99.95%.

Background: Agriculture has always been the source of global food security but is challenged by crop diseases. Beans, a vital protein source, are particularly vulnerable to the diseases whch may occur due to varous pathogens ncludng bacteria fung etc. Accurate disease identification has become critical in order to maintain increasing demands of growing population. Methods: This study utilized the AlexNet convolutional neural network (CNN) to classify bean leaf images into three classes (two disease classes i.e Angular Leaf Spot and Rust and one Healthy class). An open dataset containing leaf mages of beans belonging to all the three classes was used to train the model. The mages were first preprocessed and resized to 224x224 pixels for optimal model performance. The AlexNet model was trained for 25 epochs using cross-entropy loss, ReLU activation, max-pooling and dropout regularization. Result: An accuracy of 95.4% was achieved while the validation accuracy was 78.2%. Other performance metrics, such as precision, recall and F1-score, highlighted strengths in identifying healthy leaves, with an overall accuracy of 82.81%. However, some misclassifications occurred between disease classes due to visual similarities. The results demonstrate AlexNet’s potential for automated plant disease detection, providing a scalable solution for enhancing agricultural practices and food security. Further optimization and integration with field applications are recommended for improved accuracy and usability.

The increasing complexity of modern software development necessitates tools and methodologies for code analysis, maintenance, and migration in multi-language Integrated Development Environments (IDEs). The security needs of the software development process also recently led to the introduction of the Software Bill of Materials (SBOM), which vendors must include in their products. Automated tools become imperative to speed up software production and distribution. In the present code development landscape, automated Programming Language Identification (PLI) helps to develop secure software development with its in-depth insights into programming consequences for code maintenance, legacy system management, code analysis, quality assurance, software modernization, migration, and code search. With the increasing success of Machine Learning (ML) models in detection methods, researchers use them for PLI showing superior performance of detection compared to the basic method of identification based on the source file extension. In this paper, we present the first survey for ML-based PLI methods providing insights into the domain’s present status and guiding towards a futuristic tool development for PLI. We evaluate various ML techniques that recognize programming languages. These techniques include conventional, operation-based, and data source-based techniques. Our study examines the advantages and limitations of identifying the programming language in text, images, and videos. Our investigation also emphasizes the capabilities of the existing solutions that improve software development practices. Furthermore, our survey analyzes the strengths and limitations of existing code detection tools such as GitHub’s Linguist, Pygments, highlight.js, Ace, Google’s Code Prettify, SourcererCC, Guesslang, and SonarLint. We also provide some open research problems for future researchers.

This research explores an innovative approach to segment hyperspectral images. Aclass-aware feature-based attention approach is combined with an enhanced attention-based network, FAttNet is proposed to segment the hyperspectral images semantically. It is introduced to address challenges associated with inaccurate edge segmentation, diverse forms of target inconsistency, and suboptimal predictive efficacy encountered in traditional segmentation networks when applied to semantic segmentation tasks in hyperspectral images. First, the class-aware feature attention procedure is used to improve the extraction and processing of distinct types of semantic information. Subsequently, the spatial attention pyramid is employed in a parallel fashion to improve the correlation between spaces and extract context information from images at different scales. Finally, the segmentation results are refined using the encoder-decoder structure. It enhances precision in delineating distinct land cover patterns. The findings from the experiments demonstrate that FAttNet exhibits superior performance compared to established semantic segmentation networks commonly used. Specifically, on the GaoFen image dataset, FAttNet achieves a higher mean intersection over union (MIoU) of 77.03% and a segmentation accuracy of 87.26% surpassing the performance of the existing network.

The integration of blockchain technology into consumer electronics is transforming healthcare by enhancing data security and streamlining operations in devices like wearable health monitors, smart medical devices, and electronic health records. However, this integration raises significant privacy concerns, especially regarding potential data leaks through automated processes like smart contracts. Traditional Private Information Retrieval (PIR) methods struggle with balancing data accessibility and patient confidentiality. This creates computational inefficiencies and interoperability challenges that hinder the implementation of standardized privacy measures in blockchain. We introduce the MUlti-Dimensional pRivacy Attainment (MUDRA) framework, a novel solution addressing critical privacy challenges in the storage and retrieval of patient and healthcare service provider (HSP) data. MUDRA leverages the innovative Dual Server Ring Private Information Retrieval (DS-RPIR) protocol to securely retrieve Personal Health Information (PHI) without publishing sensitive data to the blockchain, ensuring enhanced privacy and preventing unauthorized access. By integrating the Interplanetary File System (IPFS), MUDRA preserves data confidentiality and supports time-controlled revocation, allowing patients to set time-limited access to their medical records, automatically revoking access after the specified period. The framework introduces a new metric, Privacy Attainment (PA), to quantify transaction privacy and evaluate the private retrieval of sensitive data. When implemented on the Hyperledger Fabric network, MUDRA achieves 3.5% higher throughput (59 TPS), reduced **time complexity O(nlogn), and 8 ms latency–2.25 ms less than existing solutions. These performance improvements, coupled with robust privacy features, position MUDRA as a pioneering solution for addressing privacy concerns in the rapidly evolving field of healthcare consumer electronics.

Artificial intelligence (AI) research on video games primarily focused on the imitation of human-like behavior during the past few years. Moreover, to increase the perceived worth of amusement and gratification, there is an enormous rise in the demand for intelligent agents that can imitate human players and video game characters. However, the agents developed using the majority of current approaches are perceived as rather more mechanical, which leads to frustration, and more importantly, failure in engagement. On that account, this study proposes an imitation learning framework to generate human-like behavior for more precise and accurate reproduction. To build a computational model, two learning paradigms are explored, artificial neural networks (ANN) and adaptive neuro-fuzzy inference systems (ANFIS). This study utilized several variations of ANN, including feed-forward, recurrent, extreme learning machines, and regressions, to simulate human player behavior. Furthermore, to find the ideal ANFIS, grid partitioning, subtractive clustering, and fuzzy c-means clustering are used for training. The results demonstrate that ANFIS hybrid intelligence systems trained with subtractive clustering are overall best with an average accuracy of 95%, followed by fuzzy c-means with an average accuracy of 87%. Also, the believability of the obtained AI agents is tested using two statistical methods, i.e., the Mann–Whitney U test and the cosine similarity analysis. Both methods validate that the observed behavior has been reproduced with high accuracy.

Heavy metal ions (HMIs) are very harmful to the ecosystem when they are present in excess of the recommended limits. They are carcinogenic in nature and can cause serious health issues. So, it is important to detect the metal ions quickly and accurately. The metal ions arsenic (As3+), cadmium (Cd2+), chromium (Cr3+), lead (Pb2+), and mercury (Hg2+) are considered to be very toxic among other metal ions. Standard analytical methods like atomic absorption spectroscopy, atomic fluorescence spectroscopy, and X-ray fluorescence spectroscopy are used to detect HMIs. But these methods necessitate highly technical equipment and lengthy procedures with skilled personnel. So, electrochemical sensing methods are considered to be more advantageous because of their quick analysis with precision and simplicity to operate. They can detect a wide range of heavy metals providing real-time monitoring and are cost-effective and enable multiparametric detection. Various sensing applications necessitate severe regulation regarding the modification of electrode surfaces. Numerous nanomaterials such as graphene, carbon nanotubes, and metal nanoparticles have been extensively explored as interface materials in electrode modifiers. These nanoparticles offer excellent electrical conductivity, distinctive catalytic properties, and high surface area resulting in enhanced electrochemical performance. This review examines different HMI detection methods in an aqueous medium by an electrochemical sensing approach and studies the recent developments in interface materials for altering the electrodes.

Background Feature selection is a vital process in data mining and machine learning approaches by determining which characteristics, out of the available features, are most appropriate for categorization or knowledge representation. However, the challenging task is finding a chosen subset of elements from a given set of features to represent or extract knowledge from raw data. The number of features selected should be appropriately limited and substantial to prevent results from deviating from accuracy. When it comes to the computational time cost, feature selection is crucial. A feature selection model is put out in this study to address the feature selection issue concerning multimodal. Methods In this work, a novel optimization algorithm inspired by cuckoo birds’ behavior is the Binary Reinforced Cuckoo Search Algorithm (BRCSA). In addition, we applied the proposed BRCSA-based classification approach for multimodal feature selection. The proposed method aims to select the most relevant features from multiple modalities to improve the model’s classification performance. The BRCSA algorithm is used to optimize the feature selection process, and a binary encoding scheme is employed to represent the selected features. Results The experiments are conducted on several benchmark datasets, and the results are compared with other state-of-the-art feature selection methods to evaluate the effectiveness of the proposed method. The experimental results demonstrate that the proposed BRCSA-based approach outperforms other methods in terms of classification accuracy, indicating its potential applicability in real-world applications. In specific on accuracy of classification (average), the proposed algorithm outperforms the existing methods such as DGUFS with 32%, MBOICO with 24%, MBOLF with 29%, WOASAT 22%, BGSA with 28%, HGSA 39%, FS-BGSK 37%, FS-pBGSK 42%, and BSSA 40%.

Accurate and flexible 3D pose estimation for virtual entities is a strenuous task in computer vision applications. Conventional methods struggle to capture realistic movements; thus, creative solutions that can handle the complexities of genuine avatar interactions in dynamic virtual environments are imperative. In order to tackle the problem of precise 3D pose estimation, this work introduces TRI-POSE-Net, a model intended for scenarios with limited supervision. The proposed technique, which is based on ResNet-50 and includes integrated Selective Kernel Network (SKNet) blocks, has proven to be efficient for feature extraction customised specifically to pose estimation scenarios. Furthermore, trifocal tensors and their trio-view geometry allow us to generate 3D ground truth poses from 2D poses, resulting in more refined triangulations. Through the proposed approach, the 3D poses can be estimated from a single 2D RGB image. Moreover, the proposed approach was evaluated on the HumanEva-I dataset yielding a Mean-Per-Joint-Position-Error (MPJPE) of 47.6 under self-supervision and an MPJPE of 29.9 under full supervision. In comparison with the other works, the proposed work has performed well in the self-supervision paradigm.

Vehicular networks are expanding their applications for future sustainability. The reported increasing rate of data breaches through vehicular networks by Distributed Denial of Service (DDoS) type of intrusion creates concern for such networks. Existing security solutions focus only on intrusion detection. However, prevention solutions are more proactive and provide security by probabilistic analysis. Existing prevention models for vehicular networks have low accuracy and are unable to handle zero-day attacks and advanced persistent threats. In this paper, we solve the problems mentioned above and introduce Predictive Risk Evaluation for Vehicular Infrastructure Resilience (PREVIR), the first amalgamated model of logit method (statistical analysis) and LogitBoost method (machine learning) to prevent DDoS attacks in vehicular networks. In PREVIR, the logit model predicts the packet probabilities for identifying maliciousness. The machine learning method improves PREVIR’s performance through iterative refinement of the model’s periodic updates based on new traffic parameters. We run a set of experiments on PREVIR. We use our NS3-generated dataset, NSL-KDD public dataset, and CIC-DDoS public dataset. PREVIR analyses multiple attack types, including UDP flood, TCP flood, mixed flooding, U2R, Probe, and R2L attacks. The results show that PREVIR classifies packets with accuracy up to 99.99%. Our proposed PREVIR model achieves a True Positive Ratio (TPR) up to 100% and an average False Positive Ratio (FPR) of 35%. The comparative analysis shows that PREVIR’s efficiency is 20% better on average in the prevention of malicious packets as compared to the state-of-the-art models.

Quora is an expanding online platform, that contains a growing collection of questions and answers generated by users. The content on this platform is managed by its users which involves creating, editing, and organization. Due to the vast number of users, it is not uncommon to find multiple questions with similar intents, leading to the problem of duplicate and identical questions. Detection of these duplicates could effectively lead to a more efficient search for high-quality answers, ultimately improving the user experience for both readers and writers on Quora. This study utilizes the dataset of Question Pairs for Quora obtained from Kaggle for identifying questions that are duplicates or identical. To vectorize the questions and for model training, six types of word embeddings are implemented including GoogleNewsVector, FastText crawl, FastText crawl sub-words, bidirectional encoder representations from transformers (BERT), robustly optimized BERT pretraining approach (RoBERTa), and embeddings from language models (ELMO) containing 100 dimensions. The Siamese Manhattan long short-term memory (MaLSTM) neural network model, where Ma is Manhattan distance, is applied with ELMO word embedding to predict duplicate questions in the dataset. Experimental results demonstrate that the proposed model attained an accuracy of 95.68% which surpasses the state-of-the-art models.

Biometric stress monitoring has become a critical area of research in understanding and managing health problems resulting from stress. One of the fields that emerged in this area is biometric stress monitoring, which provides continuous or real-time information about different anxiety levels among people by analyzing physiological signals and behavioral data. In this paper, we propose a new approach based on the CapsNets model for continuously monitoring psychophysiological stress. In the new model, streams of biometric data, including physiological signals and behavioral patterns, are taken up for analysis. In testing using the Swell multiclass dataset, it performed with an accuracy of 92.76%. Further testing of the WESAD dataset reveals an even better accuracy at 96.76%. The accuracy obtained for binary classification of stress and no stress class is applied to the Swell dataset, where this model obtained an outstanding accuracy of 98.52% in this study and on WESAD, 99.82%. Comparative analysis with other state-of-the-art models underlines the superior performance; it achieves better results than all of its competitors. The developed model is then rigorously subjected to 5-fold cross-validation, which proved very significant and proved that the proposed model could be effective and efficient in biometric stress monitoring.

The correctness and the true validated data in Human Resource Management (HRM) are important for organizations as the data plays an impactful role in recruiting, developing, and retaining a skilled workforce. On one hand, the validated data in an organization helps in recruiting legitimate skillful employees; on the other hand, keeping the employee's data safe and maintaining privacy laws such as compliance with the General Data Protection Regulation (GDPR) is also an organization's responsibility. Besides, transparency in human resource management operations is crucial because it promotes trust and fairness within an organization. The present HRM systems are centralized in nature and their verifiable credential system is ineffective; this leads to the intentions of internal data sabotage or internal threats. Besides, the organizations' biases also become more prominent.In this paper, we address the above-mentioned problems with a blockchain framework for HRM to utilize the privacy of data access through a Privacy Information Retrieval (PIR) process. To be specific, our proposed framework called Blockchained piR of resOurces as humaN (BRON), is the first blockchain framework to show an effective mechanism to access data from organizations globally without hampering privacy. BRON uses a generalized user registration process to use the services of data access and in the background, it uses Zero-Knowledge Proofs (ZKPs) for global verification and PIR for privacy-based data retrieval. More specifically, credential verification and ZKP-based PIR are the highlights of our proposed BRON. Another interesting aspect of BRON is the use of Proof-of-Authority (PoA) to validate the anonymity and unlinkability of any HR operation. Finally, BRON has also contributed with a smart contract to incentivize the employees. BRON is very generic and easily be customizable as per the HR requirements. We run a set of experiments on BRON and observe that it is successful in providing privacy-assured data access and decentralized human resource data management. Overall, BRON provides 30% reduced latency and 35% better throughput as compared to the existing blockchain solutions in the direction of HRM.

Mobile consumer devices have become essential components of modern society, but their reliance on rechargeable batteries creates issues with overcharging, overdischarging, and battery health. The demand for effective and privacy-aware battery management becomes even more obvious in view of the fast changing environment of Smart IoT devices. Reliable power supplies and predicted battery performance are essential for the seamless integration of these smart IoT devices into daily life. This research offers a novel recurrent neural network (RNN) based training method for battery management systems in mobile consumer electronics and smart IoT devices. This approach addresses the heterogeneous ecosystem of smart IoT devices by allowing model training across decentralized data sources while protecting user data privacy by Federated Split Learning (FSL). Sharing extensive model parameters from these devices is restricted by FSL’s emphasis on privacy. To overcome this, it is suggested to break the RNN model into smaller sub-networks, each of which would be trained using data segments from various consumer electronics products and intelligent IoT sensors. Following their transmission to a federated server for aggregation, these sub-networks help advance mobile technology, consumer electronics, and the larger Smart IoT ecosystem by producing a more reliable RNN model for battery performance prediction.

Maternal healthcare is a critical aspect of public health that focuses on the well-being of pregnant women before, during, and after childbirth. It encompasses a range of services aimed at ensuring the optimal health of both the mother and the developing fetus. During pregnancy and in the postpartum period, the mother’s health is susceptible to several complications and risks, and timely detection of such risks can play a vital role in women’s safety. This study proposes an approach to predict risks associated with maternal health. The first step of the approach involves utilizing principal component analysis (PCA) to extract significant features from the dataset. Following that, this study employs a stacked ensemble voting classifier which combines one machine learning and one deep learning model to achieve high performance. The performance of the proposed approach is compared to six machine learning algorithms and one deep learning algorithm. Two scenarios are considered for the experiments: one utilizing all features and the other using PCA features. By utilizing PCA-based features, the proposed model achieves an accuracy of 98.25%, precision of 99.17%, recall of 99.16%, and an F1 score of 99.16%. The effectiveness of the proposed model is further confirmed by comparing it to existing state of-the-art approaches.

Breast cancer has been a significant contributor to cancer-related mortality, but advancements in early detection through regular mammography and improvements in treatment modalities have contributed to declining mortality rates in several regions. This study presents a novel approach to cancer diagnosis utilizing Full-Field Digital Mammography images through predictive analysis methods. By using predictive analytic techniques and mammography images, this study offers a novel way to cancer detection. The research involves the application of deep learning techniques to extract valuable insights from cancer images captured by mammography devices. The CBIS-DDSM (Curated Breast Imaging Subset of Digital Database for Screening Mammography) dataset including images from patients with varying types and stages of cancer, is collected and pre-processed to ensure uniformity and quality. Relevant features, including color, texture, and shape characteristics, are extracted, and a rigorous feature selection process is employed to identify discriminative markers. The Residual Network (ResNet) model is selected and trained on the dataset, with a focus on classification accuracy and robust predictive performance. Validation metrics, such as accuracy, IoU (Intersection over Union) score, dice score, and ROC (Receiver Operating Characteristic) curve are employed to evaluate the model’s efficiency. After analysis, the proposed method had the best degree of mass lesion detection accuracy, at 99.24%. This research contributes to the advancement of non-invasive and efficient diagnostic tools, potentially enhancing early detection and intervention in cancer patients. The proposed method not only demonstrates promising results in terms of diagnostic accuracy but also emphasizes interpretability, seamless integration into clinical workflows, and adherence to ethical standards.

The statewide consumer transportation demand model analyzes consumers’ transportation needs and preferences within a particular state. It involves collecting and analyzing data on travel behavior, such as trip purpose, mode choice, and travel patterns, and using this information to create models that predict future travel demand. Naturalistic research, crash databases, and driving simulations have all contributed to our knowledge of how modifications to vehicle design affect road safety. This study proposes an approach named PODE that utilizes federated learning (FL) to train the deep neural network to predict the truck destination state, and in the context of origin-destination (OD) estimation, sensitive individual location information is preserved as the model is trained locally on each device. FL allows the training of our DL model across decentralized devices or servers without exchanging raw data. The primary components of this study are a customized deep neural network based on federated learning, with two clients and a server, and the key preprocessing procedures. We reduce the number of target labels from 51 to 11 for efficient learning. The proposed methodology employs two clients and one-server architecture, where the two clients train their local models using their respective data and send the model updates to the server. The server aggregates the updates and returns the global model to the clients. This architecture helps reduce the server’s computational burden and allows for distributed training. Results reveal that the PODE achieves an accuracy of 93.20% on the server side.

Recent developments and the widespread adoption of remote sensing data (RSD) gave rise to various hyperspectral imaging (HSI) applications. With its detailed spectral information, HSI has been adopted for use in various agriculture-related applications. These applications demand more accurate classification, highlighting the significance of HSI in plant disease detection, crop classification, etc. Despite existing models for RSD-based agricultural applications, such models lack generalizability for plant classification. This task is challenging for the UNet-based architectures due to nonlinear combinations of the pixels in the hyperspectral image, reduced and diverse information of each pixel, and high dimensions. Furthermore, a shortage of labeled data makes achieving high classification accuracy challenging. This study proposes an improved 3-D UNet architecture based on a modified convolutional neural network that uses spatial and spectral information. This approach solves the limitations of existing UNet-based models, which suffer to deal with nonlinear combinations and reduced and diverse information of small pixels. The proposed model employs a semantic segmentation strategy with modified architecture for more accurate classification and segmentation. The studies employ widely recognized benchmark HSI datasets, such as the Indian Pines, Salinas, Pavia University, Honghu, and Xiong'an datasets. These datasets are assessed using average accuracy, overall accuracy, and the Kappa coefficient. The suggested model demonstrated exceptional classification accuracy, achieving 99.60% for the Indian Pines dataset and 99.67% for the Pavia University dataset. The proposed approach is further validated and proven to be superior and robust through comparisons with existing models.

As organizations increasingly harness big data for analytics and decision-making, the efficient processing of massive datasets becomes paramount. Hadoop, a widely adopted distributed computing framework, excels in processing large-scale data. However, its performance is contingent on effective data locality, which becomes challenging in heterogeneous computing environments comprising diverse hardware resources. This research addresses the imperative of enhancing Hadoop’s data locality performance in heterogeneous computing environments. The study explores strategies to optimize data placement and task scheduling, considering the diverse characteristics of nodes within the infrastructure. Through a comprehensive analysis of Hadoop’s data locality algorithms and their impact on performance, this work proposes novel approaches to mitigate challenges associated with disparate hardware capabilities. Weighted Extreme Learning Machine Technique (Weighted ELM) with the Firefly Algorithm (WELM-FF) is used in the proposed work. The integration of Weighted Extreme Learning Machine (WELM) with the Firefly Algorithm holds promise for enhancing machine learning models in the context of large-scale data processing. The research employs a combination of theoretical analysis and practical experiments to evaluate the effectiveness of the proposed enhancements. Factors such as network latency, disk I/O, and CPU capabilities are taken into account to develop a holistic framework for improving data locality and, consequently, overall Hadoop performance. The findings presented in this study contribute valuable insights to the field of distributed computing, offering practical recommendations for organizations seeking to maximize the efficiency of their Hadoop deployments in heterogeneous computing environments. By addressing the intricacies of data locality, this research strives to enhance the scalability and performance of Hadoop clusters, thereby facilitating more effective utilization of big data resources.