Hyperparameter Optimization with Genetic Algorithms and XGBoost: A Step Forward in Smart Grid Fraud Detection.

This study provides a comprehensive analysis of the combination of Genetic Algorithms (GA) and XGBoost, a well-known machine-learning model. The primary emphasis lies in hyperparameter optimization for fraud detection in smart grid applications. The empirical findings demonstrate a noteworthy enhancement in the model's performance metrics following optimization, particularly emphasizing a substantial increase in accuracy from 0.82 to 0.978. The precision, recall, and AUROC metrics demonstrate a clear improvement, indicating the effectiveness of optimizing the XGBoost model for fraud detection. The findings from our study significantly contribute to the expanding field of smart grid fraud detection. These results emphasize the potential uses of advanced metaheuristic algorithms to optimize complex machine-learning models. This work showcases significant progress in enhancing the accuracy and efficiency of fraud detection systems in smart grids.

Object Counting via Group and Graph Attention Network.

Object counting, defined as the task of accurately predicting the number of objects in static images or videos, has recently attracted considerable interest. However, the unavoidable presence of background noise prevents counting performance from advancing further. To address this issue, we created a group and graph attention network (GGANet) for dense object counting. GGANet is an encoder-decoder architecture incorporating a group channel attention (GCA) module and a learnable graph attention (LGA) module. The GCA module groups the feature map into several subfeatures, each of which is assigned an attention factor through the identical channel attention. The LGA module views the feature map as a graph structure in which the different channels represent diverse feature vertices, and the responses between channels represent edges. The GCA and LGA modules jointly avoid the interference of irrelevant pixels and suppress the background noise. Experiments are conducted on four crowd-counting datasets, two vehicle-counting datasets, one remote-sensing counting dataset, and one few-shot object-counting dataset. Comparative results prove that the proposed abbr achieves superior counting performance.

Artificial Intelligence and Blockchain Enabled Smart Healthcare System for Monitoring and Detection of COVID-19 in Biomedical Images.

Millions of individuals around the world have been impacted by the ongoing coronavirus outbreak, known as the COVID-19 pandemic. Blockchain, Artificial Intelligence (AI), and other cutting-edge digital and innovative technologies have all offered promising solutions in such situations. AI provides advanced and innovative techniques for classifying and detecting symptoms caused by the coronavirus. Additionally, Blockchain may be utilised in healthcare in a variety of ways thanks to its highly open, secure standards, which permit a significant drop in healthcare costs and opens up new ways for patients to access medical services. Likewise, these techniques and solutions facilitate medical experts in the early diagnosis of diseases and later in treatments and sustaining pharmaceutical manufacturing. Therefore, in this work, a smart blockchain and AI-enabled system is presented for the healthcare sector that helps to combat the coronavirus pandemic. To further incorporate Blockchain technology, a new deep learning-based architecture is designed to identify the virus in radiological images. As a result, the developed system may offer reliable data-gathering platforms and promising security solutions, guaranteeing the high quality of COVID-19 data analytics. We created a multi-layer sequential deep learning architecture using a benchmark data set. In order to make the suggested deep learning architecture for the analysis of radiological images more understandable and interpretable, we also implemented the Gradient-weighted Class Activation Mapping (Grad-CAM) based colour visualisation approach to all of the tests. As a result, the architecture achieves a classification accuracy of 96%, thus producing excellent results.

Convolutional Neural Networks or Vision Transformers: Who Will Win the Race for Action Recognitions in Visual Data?

Understanding actions in videos remains a significant challenge in computer vision, which has been the subject of several pieces of research in the last decades. Convolutional neural networks (CNN) are a significant component of this topic and play a crucial role in the renown of Deep Learning. Inspired by the human vision system, CNN has been applied to visual data exploitation and has solved various challenges in various computer vision tasks and video/image analysis, including action recognition (AR). However, not long ago, along with the achievement of the transformer in natural language processing (NLP), it began to set new trends in vision tasks, which has created a discussion around whether the Vision Transformer models (ViT) will replace CNN in action recognition in video clips. This paper conducts this trending topic in detail, the study of CNN and Transformer for Action Recognition separately and a comparative study of the accuracy-complexity trade-off. Finally, based on the performance analysis's outcome, the question of whether CNN or Vision Transformers will win the race will be discussed.

Thermal Modeling of Polyamide 12 Powder in the Selective Laser Sintering Process Using the Discrete Element Method.

Selective laser sintering (SLS) is one of the key additive manufacturing technologies that can build any complex three-dimensional structure without the use of any special tools. Thermal modeling of this process is required to anticipate the quality of the manufactured parts by assessing the microstructure, residual stresses, and structural deformations of the finished product. This paper proposes a framework for the thermal simulation of the SLS process based on the discrete element method (DEM) and numerically generated in Python. This framework simulates a polyamide 12 (PA12) particle domain to describe the temperature evolution in this domain using simple interaction laws between the DEM particles and considering the exchange of these particles with the boundary planes. The results obtained and the comparison with the literature show that the DEM frame accurately captures the temperature distribution in the domain scanned by the laser. The effect of laser power and projection time on the temperature of PA12 particles is investigated and validated with experimental settings to show the reliability of DEM in simulating powder-based additive manufacturing processes.

A Smart-Anomaly-Detection System for Industrial Machines Based on Feature Autoencoder and Deep Learning.

Machine-health-surveillance systems are gaining popularity in industrial manufacturing systems due to the widespread availability of low-cost devices, sensors, and internet connectivity. In this regard, artificial intelligence provides valuable assistance in the form of deep learning methods to analyze and process big machine data. In diverse industrial applications, gears are considered a condemning element; many contributing failures occur due to an unexpected breakdown of the gears. In recent research, anomaly-detection and fault-diagnosis systems have been the gears' most contributing content. Thus, in work, we presented a smart deep learning-based system to detect anomalies in an industrial machine. Our system used vibrational analysis methods as a deciding tool for different machinery-maintenance decisions. We will first perform a data analysis of the gearbox data set to analyze the data's insights. By calculating and examining the machine's vibration, we aim to determine the nature and severity of the defect in the machine and hence detect the anomaly. A gearbox's vibration signal holds the fault's signature in the gears, and earlier fault detection of the gearbox is achievable by examining the vibration signal using a deep learning technique. Therefore, we aim to propose a 6-layer autoencoder-based deep learning framework for anomaly detection and fault analysis using a publically available data set of wind-turbine components. The gearbox fault-diagnosis data set is utilized for experimentation, including collecting vibration attributes recorded using SpectraQuest's gearbox fault-diagnostics simulator. Through comprehensive experiments, we have seen that the framework gains good results compared to others, with an overall accuracy of 91%.

Automated Pulmonary Nodule Classification and Detection Using Deep Learning Architectures.

Recent advancement in biomedical imaging technologies has contributed to tremendous opportunities for the health care sector and the biomedical community. However, collecting, measuring, and analyzing large volumes of health-related data like images is a laborious and time-consuming job for medical experts. Thus, in this regard, artificial intelligence applications (including machine and deep learning systems) help in the early diagnosis of various contagious/ cancerous diseases such as lung cancer. As lung or pulmonary cancer may have no apparent or clear initial symptoms, it is essential to develop and promote a Computer Aided Detection (CAD) system that can support medical experts in classifying and detecting lung nodules at early stages. Therefore, in this article, we analyze the problem of lung cancer diagnosis by classification and detecting pulmonary nodules, i.e., benign and malignant, in CT images. To achieve this objective, an automated deep learning based system is introduced for classifying and detecting lung nodules. In addition, we use novel state-of-the-art detection architectures, including, Faster-RCNN, YOLOv3, and SSD, for detection purposes. All deep learning models are evaluated using a publicly available benchmark LIDC-IDRI data set. The experimental outcomes reveal that the False Positive Rate (FPR) is reduced, and the accuracy is enhanced.

Fast Constant-Time Modular Inversion over Fp Resistant to Simple Power Analysis Attacks for IoT Applications.

The advent of the Internet of Things (IoT) has enabled millions of potential new uses for consumers and businesses. However, with these new uses emerge some of the more pronounced risks in the connected object domain. Finite fields play a crucial role in many public-key cryptographic algorithms (PKCs), which are used extensively for the security and privacy of IoT devices, consumer electronic equipment, and software systems. Given that inversion is the most sensitive and costly finite field arithmetic operation in PKCs, this paper proposes a new, fast, constant-time inverter over prime fields Fp based on the traditional Binary Extended Euclidean (BEE) algorithm. A modified BEE algorithm (MBEEA) resistant to simple power analysis attacks (SPA) is presented, and the design performance area-delay over Fp is explored. Furthermore, the BEE algorithm, modular addition, and subtraction are revisited to optimize and balance the MBEEA signal flow and resource utilization efficiency. The proposed MBEEA architecture was implemented and tested on Xilinx FPGA Virtex #5, #6, and #7 devices. Our implementation over Fp (length of p = 256 bits) with 2035 slices achieved one modular inversion in only 1.12 µs on Virtex-7. Finally, we conducted a thorough comparison and performance analysis to demonstrate that the proposed design outperforms the competing designs, i.e., has a lower area-delay product (ADP) than the reported inverters.

Computing Techniques for Environmental Research and Public Health.

Human bodies are continuously generating information about our health [...].

Two-Dimensional Permutation Vectors' (PV) Code for Optical Code Division Multiple Access Systems.

In this paper, we present a new algorithm to generate two-dimensional (2D) permutation vectors' (PV) code for incoherent optical code division multiple access (OCDMA) system to suppress multiple access interference (MAI) and system complexity. The proposed code design approach is based on wavelength-hopping time-spreading (WHTS) technique for code generation. All possible combinations of PV code sets were attained by employing all permutations of the vectors with repetition of each vector weight (W) times. Further, 2D-PV code set was constructed by combining two code sequences of the 1D-PV code. The transmitter-receiver architecture of 2D-PV code-based WHTS OCDMA system is presented. Results indicated that the 2D-PV code provides increased cardinality by eliminating phase-induced intensity noise (PIIN) effects and multiple user data can be transmitted with minimum likelihood of interference. Simulation results validated the proposed system for an agreeable bit error rate (BER) of 10-9.

Entropy-Based Algorithms for Signal Processing.

Entropy, the key factor of information theory, is one of the most important research areas in computer science [...].

Optimization of UHF RFID Five-Slotted Patch Tag Design Using PSO Algorithm for Biomedical Sensing Systems.

In this paper, a new flexible wearable radio frequency identification (RFID) five-shaped slot patch tag placed on the human arm is designed for ultra-high frequency (UHF) healthcare sensing applications. The compact proposed tag consists of a patch structure provided with five shaped slot radiators and a flexible substrate, which minimize the human body's impact on the antenna radiation performance. We have optimized our designed tag using the particle swarm optimization (PSO) method with curve fitting within MATLAB to minimize antenna parameters to achieve a good return loss and an attractive radiation performance in the operating band. The PSO-optimized tag's performance has been examined over the specific placement in some parts of the human body, such as wrist and chest, to evaluate the tag response and enable our tag antenna conception in wearable biomedical sensing applications. Finally, we have tested the robustness of this tag by evaluating its sensitivity as a function of the antenna radiator placement over the ground plane or by shaping the ground plane substrate for the tag's position from the human body. Our numerical results show an optimal tag size with good matching features and promising read ranges near the human body.

Link-quality measurement and reporting in wireless sensor networks.

Wireless Sensor networks (WSNs) are created by small hardware devices that possess the necessary functionalities to measure and exchange a variety of environmental data in their deployment setting. In this paper, we discuss the experiments in deploying a testbed as a first step towards creating a fully functional heterogeneous wireless network-based underground monitoring system. The system is mainly composed of mobile and static ZigBee nodes, which are deployed on the underground mine galleries for measuring ambient temperature. In addition, we describe the measured results of link characteristics such as received signal strength, latency and throughput for different scenarios.