Micro-Electronics and Telecommunication Engineering

The Internet of Things (IoT) refers to the network that connects objects with various processing and sensing capabilities and communicates through the Internet. IoT facilitates connectivity between numerous technological gadgets and sensors to improve our quality of life. It has revolutionized the world around us. IoT has a great impact on transportation as IoT integration allows in growing a centrally managed network that may optimize the space protected via way of means of the vehicle, locate higher and secure routes in case of an important situation, effectively manipulate and keep the goods, fabric, and buy orders and universal offers a fine effect at the sales generated via way of means of the transportation sector. This paper briefly discusses the basics of IoT, the various technologies that comprise IoT, its integration with transportation, and its applications. A vertical handoff scheme is proposed for transportation and it is compared with various states of art techniques. It is observed that it outperforms over the other techniques in terms of throughput and number of end-users served.

Due to its potential applications in domains involving psychology, social smart machines, and human–computer interaction, speech emotion recognition has become an important and growing topic field in the study. That article offers the service comparative analysis of strategies in artificial intelligence classifiers for speech emotion recognition, including MLP, decision tree, SVM, and random forest. Using visual analysis methods like wave plot and spectrogram analysis as well as audio feature extraction, we evaluate these classifiers. According to our observations, MLP achieves better performance than the other classifiers, obtaining an accuracy of 85% when identifying different emotions. Moreover, we illustrate how audio feature extraction and visual analysis help to improve emotion recognition efficiency. Our research has applications for creating speech emotion recognition algorithms that may be applied in real-life scenarios.

Due to its sharp subthreshold swing and low leakage current, VDL-TFET has become a potential option for low-power electronic devices. In this study, we examine the influence of various device settings on the electrical properties of VDL-TFETs to assess their performance in low-power applications. We analysis how the doping level, layer thickness, gate length, and temperatures affect the functionality of VDL-TFETs. Simulations conducted by us demonstrate that by carefully tuning those parameters, the device’s ON-current, subthreshold swing, and delay properties may be greatly enhanced, rendering it appropriate for low-power digital and analogue electronics. We additionally contrast the efficiency of VDL-TFETs against that of various other low-power transistors, including FinFETs, and show that the subthreshold swing and delay features of VDL-TFETs are comparable to those of FinFETs. According to our research, VDL-TFETs are an intriguing technology for low-power semiconductor purposes.

The goal of this work is to create an emotional model that can categorize emotions in real-time for persons with Asperger’s syndrome (AS). The model is based on facial expressions, head movement, and eye gaze as significant features for emotions. Developing like this an emotional model can aid other people to communicate socially with people with autism. This model overcomes the limitations of previous research that used sensitive and invasive methods to capture physiological data in order to predict emotional states. These instruments are very costly and need a controlled environment. The proposed model succeeded in classifying the emotional states of the AS using a natural spontaneous dataset without invasive tools and in an uncontrolled environment. The dataset used in this study is an available dataset and contains videos recorder in an uncontrolled environment with different facial occlusion and illumination changes. The emotions were classified as fear, disgust, joy, anticipation, and sadness. The proposed model implements a modified version of the well-known and wide-used GoogleNet machine learning model to classify emotions. The key feature of GoogleNet is the inception module, which is designed to capture different levels of spatial features and perform dimensionality reduction using parallel convolutional layers with different filter sizes. There are several metrics that were used to measure the accuracy and generality of a DL model, including accuracy, precision, recall, and F1 score, depending on the nature of the task. Overall, measuring the generality of a DL model is an important step in evaluating its performance and ensuring that it will perform well on new, unseen data. The model achieved significant performance on unseen data with accuracy (98%).

The amount of automobiles on the road is rising dramatically as well as the commercial revolution and the economy increase. This work seeks to recognize and categorize significant objects in a video stream (surveillance video); moreover, it is desirable to detect and recognize information on four-wheeler license plates. Furthermore, we examine the dataset and discuss the findings. Visualizations and conclusions are also included in the discussion. The performance of several classifiers is then compared using object classification. The accuracy of the model is investigated as a function of data multiplication. Finally, we will go through the methods and strategies utilized to recognize license plates. ALPR makes use of image processing for recognizing vehicles by their license plates without the direct involvement of a person by extracting the information from the picture of the vehicle or from a series of images. In this study, CNN, a deep learning approach, is used to distinguish as well as identify car license plates. This technology can recognize and locate license plate numbers. Digital cameras are used to detect vehicles.

Car manufacturing rates have climbed consistently over the previous decade, with million cars manufactured per year. This has given a significant boost to the vintage car market, whose is now entrenched as a thriving industry. The recent entrance of many internet portals and platforms has supplied buyers, customers, merchants, and vendors with the necessity to keep informed with the current situation and trends in order to know the true worth of a used automobile in the contemporary market. While there are several uses of machine learning in realistic life, one of their most apparent aspects is its usage in addressing prediction issues. Furthermore, there are an infinite variety of topics on which predictions can become made. This paper is about and revolves around a particular application. We will anticipate the market value of an obsolete vehicle using a machine learning method such as linear regression and construct a mathematical structure utilizing data that has been provided with a certain set of attributes.

This work presents an evanescent wave-based U-shaped plastic optical fiber sensor. The effect of bend-induced material deformation on numerical aperture (N.A) and V-number has been thoroughly investigated. With sufficiently smaller bend radius, the numerical aperture of a U-bent fiber decreases toward zero near the inner curvature of the bending. The present scenario causes entire optical power loss at interface between core-clad (sample region) interfaces, resulting in no detectable power. It has been observed that as the bending radius increases, the local N.A value decreases while the refractive index of the surrounding region remains fixed. As a result, the fractional power in the surrounding region rises, resulting in increased absorbance and sensitivity of the EW absorption-based optical fiber sensor.

High blood sugar levels in diabetes mellitus (DM) can cause cardiac arrest, nervous system damage, vision loss, foot problems, liver or kidney damage, and death if left untreated. Age, gender, family history, BMI, and glucose levels all contribute to diabetes. To increase diabetes detection and prevent health concerns, machine learning techniques are used for prediction. Identifying the type of diabetes and considering the risk of accompanying diseases can improve diabetes prediction accuracy. This study uses one-way analysis of variance, mutual information, and F-regressor with random forest, Gaussian Naive Bayes, support vector machine, and decision tree for feature selection. Results with and without selected algorithms are compared. They have been used to adjust diabetic care using clinical parameters like accuracy, precision, recall, and F1-score. Random forest (RF) using F-regressor (FR) or ANOVA feature selection and numerous iterations of N (75) and K (3–5) outperforms competitors with 0.9 accuracy. This proves the diabetes-related DNA sequence classification technique works.

Financial fraud is the illegal use of mobile platforms for transactions when credit card or identity theft is exploited to create fake money. With the spread of smartphones and online transaction services, financial fraud and credit card fraud have become rapidly growing issues. Accurate detection of financial fraud in this context is crucial, as it can result in significant financial losses. Therefore, we conducted a survey of machine learning, deep learning, and data mining methodologies for financial fraud detection. In our study, we evaluated our methodology for detecting fraud and handling vast financial data, contrasting it with artificial neural networks. Our reviewed process encompassed variable selection, sampling, and the utilisation of supervised and unsupervised algorithms. This allowed us to effectively identify financial fraud and process extensive financial datasets.

While there has been a massive growth in online shopping platforms. There data has been mostly stored in centralized storages which mean they can be an easy target to fraud and theft. It is important to preserve the data of the product and the user in a secure and decentralized manner. With the advent of blockchain technology across the horizon, it is necessary to build a marketplace model that works in a reliable way on this technology while combating the shortcomings of the previous models. Aim of this paper is to address the issues and propose a model that can get reduced fees costs while uploading the data to the blockchain that is uploading the item for sale and while purchasing the item, we also as well as find a solution to collusion from different parties. The end result of this paper should be a fully functional application that is capable of reducing gas fees and increases user reliability on the platform.

Wildfires are a major natural disaster that can cause significant damage to ecosystems and human communities. Wildfire behavior must be predicted accurately for effective emergency response and evacuation planning. The proposed system suggests a novel Convolutional Neural Networks (CNNs) method to estimate wildfire conditions via optimization of perimeter and surface area. Extraction of the required features is from the historical wildfire data which is performed before preprocessing and training. The trained CNN is then validated and optimized for performance, with the goal of accurately predicting wildfire behavior in real-time. The proposed system results show the effectiveness of the method, improving the wildfire prediction ability. The outcome of the prediction determines the probability of a wildfire. The results can be used to monitor areas where wildfires are expected. This helps to implement strategies that can be used to mitigate the effects of wildfires. The combination of perimeter and surface area optimization with CNNs represents a promising new approach to wildfire prediction and management, with broad applications in land management, sustainability, and emergency response.

Cloud-based data storage has become ubiquitous in e-commerce, education, research, and health care. Among these industries, health care has the potential to leverage cloud technology to store electronic health records (EHRs) and reduce costs associated with maintaining manual paper-based records. The storage, retrieval, and maintenance of electronic health records for healthcare providers have been challenging due to the influx of patients and the requirement to retain their records for extended periods. Consequently, many healthcare providers are keenly interested in transitioning their EHRs to cloud-based platforms. The risks mentioned above stem from storing electronic health records in distant geographical locations that may not be familiar to healthcare providers. Despite the existing proposals in the literature to address security concerns, there remains a demand for a proficient security framework capable of managing all security issues associated with using cloud storage for EHRs. The present study introduces a security framework that incorporates mechanisms to address confidentiality, integrity, and access control. The framework ensures secure storage and retrieval of electronic health records (EHRs) in cloud-based environments. The research makes a significant contribution by proposing a confidentiality mechanism named Segregation and Preserve Privacy of Sensitive Data in Cloud Storage (SPPSICS). This encryption technique is specifically designed to safeguard the privacy of sensitive data in electronic health records (EHRs) that are stored in cloud storage. The proposed mechanism endeavors to partition the sensitive and insensitive attributes of electronic health records (EHRs) and employ strong encryption techniques to safeguard the sensitive attributes.

Deep neural networks (DNNs) have shown great potential in diagnosing brain tumor disorder, but their decision-making processes can be difficult to interpret, leading to concerns about their reliability and safety. This paper presents overview of explainable artificial intelligence techniques which have been developed to improve the interpretability and transparency of DNNs and have been applied to diagnostic systems for such disorders. Based on the utilized framework of explainable artificial intelligence (XAI) in collaboration with deep learning models, authors diagnosed brain tumor with the help of convolutional neural network and interpreted its outcomes with the help of numerical gradient-weighted class activation mapping (numGrad-CAM-CNN), therefore achieved highest accuracy of 97.11%. Thus, XAI can help healthcare professionals in understanding how a DNN arrived at a diagnosis, providing insights into the reasoning and decision-making processes of the model. XAI techniques can also help to identify biases in the data used to train the model and address potential ethical concerns. However, challenges remain in implementing XAI techniques in diagnostic systems, including the need for large, diverse datasets, and the development of user-friendly interfaces. Despite these challenges, the potential benefits for improving patient outcomes and increasing trust in AI-based medical systems make it a promising area of research.

Nowadays, it has become harder to transfer money overseas. Overseas transactions have intermediaries and lengthy settlement times, leading to increased remittance taxes due to financial institutions and regulatory requirements. Based on the review, we propose a novel approach using blockchain technology for overseas transactions. By using a decentralized finance application in blockchain technology, we are creating an application that converts the fiat currency which is an actual currency into a stablecoin (USDC). This stablecoin is transferred using blockchain and the user receives it back as a fiat currency. We process overseas transactions with lower remittance taxes and in less duration efficiently and securely.

This review article discusses using Generative Adversarial Networks (GANs) for image denoising in medical images. GANs have produced optimistic results in enhancing the quality of noisy medical images, which is crucial for accurate diagnosis and treatment. The article provides an overview of the challenges in medical image denoising and the working principle of GANs. The review also summarizes the recent research on using GANs for medical image denoising and compares their performance and significance. Finally, the article discusses the future directions for GAN-based medical image denoising and its potential impact on the healthcare industry.

In this study, we describe a computer-based technique for identifying osteoporosis by analyzing medical X-ray images utilizing deep convolutional neural networks (DCNNs). During the preprocessing phase, the suggested system prepares the original picture by acquiring the area of interest, enhancing contrast, and reducing noise. Subsequently, the smudging procedure was used to improve the system's accuracy and decrease mistake by creating a nearly identical fragile region throughout the database photos. The next step is using the suggested DCNN model to diagnose the problem. To do this, the dataset was preprocessed, smudged, and then put into the model in two parts: 75% for training and 25% for testing. With Dataset 1 and Dataset 2, the diagnoses’ accuracy was 94.7% and 91.5, respectively. It is important to note that two datasets were used: Dataset 1 is the Osteoporosis Knee X-ray Dataset from Kaggle, which has two classes (osteopenia and osteoporosis), and Dataset 2 is from Mendeley, which contains three classes (osteopenia, normal, and osteoporosis).

Vibration is one of the main causes responsible for the failure of any machinery, which makes it a prominent cause of failure in the detection and prediction of a machine’s working baseline. Prediction and maintenance at the right time play a very important role in curbing hazardous situations. It also provides the required data to construct a roadmap to proceed in case of a fault. It also helps in selecting a favorable maintenance procedure, approximation of time required, and the price of maintenance or substitution of the required components. The focus of this paper is to investigate the vibration signals in a “bearing rotation mechanism” for hardware fault prediction and condition-based maintenance. This paper is based on the study and simulation of vibration signals and the primary data is taken from a test setup rig of a rotating bearing mechanism. The data is acquired through the combination of a vibration sensor and a NI DAQ (data acquisition) card to make a suitable dataset and also simulate that data with the help of MATLAB software. This study covers various parameters to study the behavior of the bearing such as Kurtosis, RMS factor, and Skewness. After considering the lack of existing datasets for the parameters necessary to make predictions for the bearing status, a new dataset DBCM or “Dataset for Bearing Condition Monitoring” was designed, containing data in terms of voltage within an interval of 0.04 s. The simulation results verified the readings obtained by hardware setup.

The bearing serves as a crucial element of any machinery with a gearbox. It is essential to diagnose bearing faults effectively to ensure the machinery’s safety and normal operation. Therefore, the identification and assessment of mechanical faults in bearings are extremely significant for ensuring reliable machinery operation. This comparative study shows the performance of fault diagnosis of bearings by utilizing various machine learning methodologies, including SVM, KNN, Linear Regression, Ridge Regression, XGB Regressor, AdaBoost Regressor, and Cat Boosting Regressor. Bearings are like the unsung heroes of the mechanical world, immensely supporting and guiding the smooth motion in everything, from your car’s wheel to the propeller in a ship. However, like other mechanical components, over the course of time, the constant use of bearings can lead to wear and tear, which may ultimately result in a fault. Bearing faults can manifest in several ways, including vibration, noise, heat, and changes in lubrication that reduce the efficiency of a machine. Therefore, it is essential to regularly monitor the bearings and inspect them to detect any issues early on. The aim of this present work is to use the various ML methodology, and their application on the bearing’s data to watch the condition of the machine’s bearing. The present work is carried out in four phases. In the first phase, the data of various loads is collected. In the second phase, the data undergoes an Exploratory Data Analysis (EDA). During the third phase, the data undergoes both training and testing processes to evaluate its effectiveness. In the fourth and final phase, the model that gives the highest accuracy among all is chosen. The present approach is based on the various machine learning algorithms and their application.

This paper introduces an image steganography system that hides a high capacity of confidential data by utilizing a distributed computing scheme that offers the embedding of smaller, equally segmented confidential images across multiple cover images. The proposed strategy developed a strong security mechanism associated with self-synchronized methods by utilizing Shamir’s Secret Sharing (SSS) Scheme. To ensure extraordinary security for hidden confidential data within the cover image, double-layer security has been designed. In the primary layer, the Secret Distributing Scheme (SDS) divides a single larger secret image into identically smaller sub-images, and the SSS scheme then creates an encoded structure of smaller sub-images for assigning and restoring the distributed decomposed confidential images within the proposed shareholders. The SSS scheme will reorder the distributed confidential decomposed images into a predefined order by the authorized recipient at the moment of revealing secrets by interpolating polynomials. The next layer is steganography, where the image is segmented into 2 × 2 squares of pixels and navigated in a crisscross way; the pixel-value difference is computed among non-overlapping Red, Green, Blue (RGB) pixels in a diagonal direction within a focused square of pixel area. Secret bits are hidden within the RGB pixels of the cover image by using the presented innovative pixel-value differencing (PVD) strategy. The results of the research indicate that the presented system provides an inventive mechanism in the form of a double layer of security and vastly enhances embedding capability.

T-Shaped Dual-Gate TFET channel provides larger tunneling junction area which offers more electrons, which enhances ION current. This paper will provide a description and see different variations by using substrate material like Silicon, Gallium Arsenide, Germanium at different channel lengths (46 nm, 36 nm, 26 nm, 16 nm, 10 nm). For insulating material, we have taken Hafnium oxides (HfO2) and Silicon-oxide. We have analyzed and compared crucial characteristics such as the ION/IOFF ratio, the subthreshold swing (SSavg), transconductance, BTBT, electric field (EF), and surface potential. Germanium material which shows better ION/IOFF ratio and better subthreshold swing, Sensitivity than Silicon and Gallium Arsenide. Electric field and Surface Potential of device getting improved as we decrease the channel width.

Proteins play a major role in determining many characteristics and functions of living beings. Prediction of protein classes and subclasses is one of the prominent topics of research in bioinformatics. Machine learning methods are widely used for prediction purposes, also applied for classification and subclassification of proteins. The problem is to classify the proteins to the corresponding subclass they belong to and choose a suitable machine learning method which can be used for better subclass classification. The objective is to compare the performances of three existing machine learning methods: logistic regression, support vector machine (SVM), and random forest, for protein subclassification. For this study the methods are implemented, and their results are compared by varying the number of samples of different subclasses and varying the number of subclasses. Logistic regression and support vector machine are used as a binary classifier for predicting multiple classes with $$log_2(n)$$ l o g 2 ( n ) number of classifiers for n class labels. It is observed that both random forest and support vector machine provide almost same accuracy for smaller data size, but as the data size increases random forest performs better than SVM.

When iris images are collected under optimal circumstances, traditional iris segmentation algorithms provide accurate findings. However, an iris identification system’s success is heavily dependent on the precision with which it segments iris pictures, particularly when dealing with irises that are non-cooperative. This research investigates the challenge of recognizing irises in low-quality photos taken under challenging lighting and other imaging situations. In order to reduce processing time and eliminate noise caused by eyelashes and eyelids, the system first acquires an iris image, then improves the image quality, detects the iris boundary and the pupil, detects the eyelids, removes the eyelashes and the shadows, and converts the iris coordinates from Cartesian to polar coordinates. The iris's characteristics are extracted with the use of a Gabor filter and then compared with the help of Euclidean distance. After using the suggested technique, we compared the outcomes with those found in the literature and found that the proposed method yields significant improvements in segmentation accuracy and recognition performance.

Reproduction of abnormal tissues within the brain due to any damage can cause major concerns for an individuals’ health which can be identified by radiologists after examining cell structure of brain that clarifies whether it belongs to benign, i.e., non-cancerous or malign, i.e., cancerous. Although it cannot be treated properly, identifying abnormal growth of tissue at very initial phase can definitely help in preventing from major issues. Most of the researchers described automated brain tumor diagnosing methods in their publications which also received the most attention to provide significant contribution in the healthcare. Authors achieved the highest accuracy of 93.72% via deploying deep learning-based models while predicting brain tumor disease as these models have ability to analyze vast amount of data and able to extract significant features accurately and efficiently as compared to the existing approaches in short duration to provide improved patient outcomes and timely treatment in the healthcare.

Apple is one of the most popular fruits all over the world, and it is also very liable to diseases like scabs, apple rot, and leaf blotch. These diseases majorly destroy the quality and lead to less healthy production; it is difficult to identify diseases in apples as they appear for a short interval of time so the most prominent way to identify infection is through the condition of their leaves. Most of the leaves are infected by scab, rust, bacteria, and viruses. Early detection is complex for farmers as they all appear the same in shape, color, and texture. Deep learning technology is contributing greatly in this area, addressing this we have proposed an accurately improved Segmentation with the CNN model using a transfer learning model, i.e., DenseNet121 with the weight of ImageNet, and by adding an extra top layer for accurate results. This study also includes the comparative analysis of Seg+ DenseNet121 and some integrated machine learning models. This experiment achieves 99.06% accuracy.

The digital revolution poses a significant challenge in terms of authenticating digital images due to the ease of image manipulation. In recent years, ensuring the validity of digital photographs has become a critical subject of research. Various watermarking methods have been devised to tackle this issue, tailored to specific applications. Nonetheless, creating a watermarking system that is both reliable and secure presents a formidable task. This article delves into the intricacies of common watermarking systems, offering comprehensive frameworks. Additionally, it presents a compilation of commonly employed specifications when designing watermarking methods for diverse purposes. The paper also explores the latest advancements in digital picture watermarking technologies, examining their strengths and weaknesses. Furthermore, it sheds light on potential future attacks employing conventional methods.

Recent developments in the field of emerging technologies including Artificial Intelligence and Machine Learning have led to wide interest in designing and developing innovative solutions in the area of traffic management and human safety. Multiple researchers have used these technologies to propose solutions for traffic sign image management that can lead to driver safety and reduction in number of accidents [10, 12, 14]. To prevent road accidents, traffic signages on roads are vital parameters that can help drivers to take timely decisions and preventive measures to avoid accidents. There is a strong need to improve the traffic sign identification and detection so that irrespective of weather conditions and degradation of sign boards, still driver navigation system is able to identify the correct signs and help in decision making ([13]; Bhatt and Tiwari, Smart traffic sign boards (STSB) for smart cities [Bhatt DP, Tiwari M (2019) Smart traffic sign boards (STSB) for smart cities. In: 2nd Smart Cities Symposium (SCS 2019), pp 1–4, March. IET]). Researchers have been using public traffic sign datasets to find ways to enhance the precision of image recognition approaches. In the previous published research, Vashisht and Kumar [21], have proposed a 3D color texture-based approach for detecting the traffic sign images by making use of ML algorithms and ANN. In this paper, the research is further extended using dimensionality reduction techniques used for feature reduction on Mapillary traffic sign image dataset. In terms of organizing the rest of the paper, the next section of background covers the previous research work done in dimensionality reduction. Next, the authors have explained the proposed methodology for feature selection and ANN design. Then, results from the implementation using ranking algorithms, classifier algorithms, and their comparisons are described. In the end, authors concluded the paper with suggested future direction of work.

Over the past few years, e-commerce has led to a significant shift in business activities from traditional methods to online platforms. Nowadays, consumers heavily rely on online reviews to guide their purchasing decisions, prompting businesses to adapt to this new reality. However, fake reviews pose a critical challenge in online reviews. Fake reviews can have serious consequences, including misleading customers and damaging the reputation of organizations. To tackle this issue, various approaches, such as natural language processing, machine learning, and sentiment analysis, have been proposed as potential solutions for detecting fake reviews. These strategies typically involve analyzing the content of reviews along with associated metadata, such as the language used, review timing, and ratings. However, differentiating between fake and genuine reviews can be challenging, as fake reviewers often employ tactics to make their reviews appear more legitimate. Despite the complexities involved, significant progress has been made in developing effective strategies for detecting fake reviews. These techniques play a crucial role in ensuring that consumers can make informed decisions based on trustworthy information while safeguarding online review systems' integrity. The main focus of this paper is to combine textual elements with other related behavioral parameters, which leads to a higher rate of perception and detection compared to other existing methods. By incorporating new behavioral variables, the proposed model enhances the accuracy of detecting fake reviews. The Elmo Model is utilized for encoding result vectors and reducing computational overhead, while the VADER model helps to determine the polarity of review text, enabling individual reviews to be evaluated. To classify real-time reviews in the Yelp dataset, a stack model is constructed using the Multinomial Naive Bayes method (MNA) and the Gradient Boosting Classifier. Compared to similar studies, the proposed model demonstrates exceptional accuracy, achieving an AUC of 82% and overall accuracy of around 98%. Overall, the integration of textual components with behavioral parameters in the proposed model offers a promising approach to effectively detect fake reviews and improve the authenticity of online review systems.

Industrial Internet of Things (IIoT) is a major element for industrial systems future. Since IIoT applications are the natural evolution of the IoT, cybersecurity is the main consideration related to IIoT adoption. Because of the similar basic architecture, IIoT inherits some security challenges from IoT. IIoT is not standing alone application. It depends on other applications such as: Cloud network and Big Data. In Cloud, the devices can be connected with each other at anytime from anywhere. Big Data application plays a very important role in storing, analyzing, security, and safety of data. This research aims to design a data governance framework for IIoT. Internet of Things (IoT) governance data security is depended on the good governance of the Cloud Computing and Big Data governance. IIoT security is very important because devices are interconnected, and if there is any hack for one of the devices, it will transfer the problem to the rest of the connected devices. The final design is to use the Encryption algorithm such as Data Encryption Standard (DES) to secure the communications between the devices. Also, the authentication algorithm (two-factor method) has been used to prove the identity of the devices and to mitigate the communication between IIoT devices. Big Data that transferred between devices may be stored or retrieved from the data centers (Cloud) to solve the storage limitation problem of the devices. The results of the practical implementation show an improvement in the data security, integrity, and accuracy. The design proves more safety in dealing with devices and in executing their duties.

In countries like India, shortage of water is a serious concern, especially in the southern states such as Tamil Nadu, Kerala and Andhra Pradesh. This shortage of water becomes entangled when there is a loss of water during transmission process. Hence, there must be innovative methodologies for water management and automation for commercial buildings in order to sort these concerns. The proposed system in this paper will execute water management in the form of planning, distributing and maintaining resources of water with IoT. The autonomous water controller includes a microcontroller (ESP8266), relay-controlled motor, a waterproof ultrasonic sensor and a float switch. The water level control system is implemented with a web dashboard arrangement for appropriate display. The presented system will be an important approach for managing water resources both for residential and commercial purposes effectively.

Cloud computing is a distributed computing architectural model that offers on-demand computing services via the Internet. It includes servers, computing resources, applications, data storage, development tools, etc. However, despite the different services and benefits offered by the cloud, security and privacy are the primary hindrances for organizations for opting cloud. The real benefit of the cloud can be only enjoyed if the sensitive data kept in the database can be secured from unauthorized access. Privacy preservation in cloud computing is nothing but hiding sensitive data, where the data is stored in the software and databases scattered around the Internet. Various methods have been put forward by researchers to preserve privacy in cloud computing over the past few years.

IoT refers to the interconnection of electronic devices, machines, and physical objects in our environment. Heterogeneous wireless sensor networks (HWSNs) are among the promising wireless technologies that play an important role in monitoring remote areas. The clustering algorithm reduces energy consumption by using a key technique. It increases the network's scalability and lifetime. Wireless sensor networks with heterogeneous characteristics should be designed with energy-efficient clustering protocols. In this paper, design a novel EEECT-IOT-HWSN technique for the three-tier heterogeneous networks. The EEECT-IOT-HWSN technique has used the modified threshold formula for the cluster head selection based on the combination of the energy and distance of the SNs. The performance of the proposed model shows the higher residual energy, less dead SNs, and higher network lifetime when compared with the ADV-LEACH1 (HETRO), and ADV-LEACH1 (HOMO) technique.

A number of fire accidents in forests occur around the globe every year which amount to catastrophes beyond all sorts of comprehensions. Behind this, many houses and lot of trees pose a serious threat to forests grown in an ambient and healthy environment. This paper enunciates an innovative methodology for detection of fires in forests using the concept of Internet of Things (IoT). IoT has gained a lot of attention due to latest advancements in technology. This work is capable of detecting forest fires in early manner by implementing a smart system integrated with temperature, moisture and humidity sensors. The system employs IoT for transmission of data to the authorized person over internet thereby increasing the efficiency of the fire detection process. The detection is intimated by means of an alarm in that specific location. The data is validated using a sensor threshold value thereby achieving maximum reliability for prevention.

The majority of people in the current atmospheric conditions are affected by lung cancer disease. The analysis of respiratory illness offers a captivating and dynamic research space with far-reaching implications for human health. A diagnostics like this can only assist in reducing the likelihood of obtaining human life in jeopardy by initial detection of metastatic disease to address this problem. Lung cancer is the leading cause of cancer death worldwide, so different algorithms have been used to forecast the prognosis of cancer patients. Because of this, patients with lung cancer are living longer on average. When making predictions, the logistic regression assessment method is more accurate than that of other methods. This report examines two additional different approaches to machine learning for forecasting a lung participant's life expectancy, including linear discriminant analysis (LDA), random forest (RF), and artificial neural networks (ANN). In order to increase success rates, various algorithms were tested. The primary goal of this is to evaluate the accuracy of classification methodologies to develop a melanoma statistical method and a resilience analysis. The correctness, accuracy, recall, and selectivity of the numerous models’ performances are assessed and compared. In this enquiry, linear discriminant analysis will perform the best among the three algorithms.

Since the advent of the Internet and digital technology, every organization uses digital tools to conduct its daily business. Finding new threats and vulnerabilities and calculating their influence on an organization's operations are the key goals of cybersecurity. To have strong cybersecurity in place, it is important to have some kind of mechanism to measure it. Metrics are used to measure something that has no direct method/instrument to measure. Cybersecurity metrics refer to the quantitative and qualitative measurements used to assess the effectiveness, efficiency, and overall health of an organization's cybersecurity efforts. By leveraging multi-criteria decision-making (MCDM) techniques in cybersecurity, organizations can improve their decision-making processes, optimize resource allocation, and enhance their overall cybersecurity posture to better defend against evolving cyber threats. Therefore, the purpose of this study is to model cybersecurity metrics evaluation by developing a decision network using analytical network process (ANP). The identification of various criteria is very important. Security metrics are organized into five primary classes divided down into 15 sub-classes, and these 15 sub-classes are further subdivided into 29 sub-classes.

Aquaculture is rapidly growing to meet the global demand for fish consumption, but it faces several challenges that adversely affect marine environments. To address these issues, economical remote-surveillance system has been developed for marine net pens, utilizing GSM, GPS modules with IoT and embedded systems technology. The goal is to create an actual time, accurate remote-surveillance system that can replace the need for frequent manual inspections by aquaculture farm workers. To overcome this, the project aims to remotely investigate aquaculture parameters such as seawater temperature, humidity, turbidity, chlorophyll, under water pressure, dissolved oxygen (DO), and pH every day, without physical visits to the pens. Ultimately, the system's objective is to enable remote control and observation of sustaining and maritime operations. The key components of the monitoring system include the fish farming enclosure monitoring station and the remote-surveillance centers. This monitoring station is equipped with intelligent water sensors to detect seawater temperature, turbidity, DO, and pH in real time. This data is collected by a remote data-collection terminal and transferred by GPRS network for online processing. The remote-surveillance centers serve as centralized servers and connect to the monitoring stations through the Internet. The seawater data collected by the particular monitoring stations is transmitted to these centers via GPRS and GSM wireless networks with the specific locations by the GPS technology, facilitating continuous monitoring and reporting. One significant challenge faced by Open-Pen Sea Cage Aquaculture is the use of fishmeal sourced from wild fisheries in the feed given to farmed fish. While alternative proteins from land-based crops are included in the feed pellets, carnivorous or omnivorous species still require some fishmeal to obtain essential amino acids. Small, oily fish such as anchovies, sardines, and pilchards, known as “Industrial Fish,” are commonly used for fishmeal. Roughly, 20 million tons of these fish are caught yearly for aquaculture, with much of Australia's sardine catch, unfit for human consumption, feeding Southern Bluefin Tuna in Port Lincoln. Relying on wild-caught fish for fishmeal raises sustainability and ecosystem impact concerns.

In today’s competitive business landscape, optimizing customer relations is paramount for sustained success. Harnessing the power of machine learning, this research presents a data-driven approach to achieve this objective. By leveraging three prominent algorithms, namely Linear Regression (LR), decision tree (DT), and support vector machine (SVM), customer behavior patterns are identified and analyzed. Through the systematic examination of vast datasets, this study attains an impressive accuracy of 95%. The findings showcase the potential of machine learning in enhancing customer relations, enabling businesses to make more informed decisions, tailor personalized experiences, and foster long-lasting customer loyalty. This data-driven approach promises to revolutionize CRM strategies, propelling enterprises toward unparalleled growth and success.

Over the past few decades, technological advancements have significantly impacted Education, transforming traditional teaching, and learning methods. Extended Reality (XR) is one of the most promising innovations in this area. XR is a buzzword that includes Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR). These technologies offer immersive and interactive experiences that enable users to interact with Virtual objects and environments, blurring the lines between the physical and digital worlds. The concept of the Metaverse has recently emerged within XR, and it has the potential to transform the Education industry. It defines a shared Virtual universe where users can collaborate, interact, and engage in various activities using Avatars. This paper uses the EON-XR platform to explore the transformative impact of the Metaverse in the Education industry.

Heart disease is the main cause of mortality across all age groups in the modern world. Thus, the necessity for improving heart attack prediction utilizing various machine learning (ML) or deep learning (DL) approaches is necessary for the health industry. Globally, the prognosis of heart disease can be improved by early diagnosis and treatment. The IoT purpose is to make simple way of making energy, wealth, and saving time easy with smart environment. The machine learning (ML) or deep learning (DL) techniques are used in various Internet of Things (IoT)-based technologies to reduce time, money, energy, and others for better performance or development. In this paper, we are going to see different kinds of machine learning and deep learning used in Internet of Things for heart disease detection system. As a starting point, we provide an overview of machine learning before moving on to explore various learning methods including deep learning models. We used EfficientNet-B0, a new convolutional network with faster training speed and better parameter efficiency than previous models. EfficientNet-B0 shows promising results for heart disease prediction.

Recent studies have shown that deep learning does pretty well at reproducing 3D scenes using multiple-view images or videos. Nevertheless, these restorations do not expose the personalities of the items, and item identification is necessary for such a scene to work in augmented worlds or interactive features. The objects in a picture that have been reconstructed as a unified mesh are handled as a singular body rather than being seen as independent categories that can be engaged with or changed. Reconstructing an entity three-dimensional image from a two-dimensional image is challenging since the transformation from a visual scene to a picture is permanent and reduces a dimensionality. In addition to creating more exact shapes when compared with previous methodologies for mesh rebuilding from individual images, our approach exhibited improved achievement in initiating comprehensive meshes when compared to strategies using only inherent portrayal mesh rebuilding networks (for instance, neighborhood deep implicit functions). By applying the multi-modal instructional methods, this was done. Real-world facts were utilized to assess the effectiveness of the suggested method. The results showed that it might perform noticeably better than earlier methods for entity 3D scene rebuilding.

The successful execution of Configure, Price, Quote (CPQ) protocols is of utmost importance for businesses that handle intricate product portfolios within their sales processes. The objective of these approaches is to optimise the generation of competitive quotations by leveraging data from diverse business systems, leading to a decrease in processing time and enhancement of operational efficiency. Salesforce's Industries CPQ for Communications Cloud, previously referred to as Vlocity CPQ, assumes a crucial function within the Salesforce platform by facilitating the automation of the quotation creation process for sales teams. Designed specifically for the communications industry, this solution enables communications service providers (CSPs) to effectively provide a wide array of products and services to their customers in a streamlined manner. By implementing Industries CPQ, Communication Service Providers (CSPs) have the potential to save expenses associated with customisation and maintenance, while also accelerating their time-to-market. In order to assess the practical ramifications of Salesforce Industries CPQ, the researcher intends to deploy the aforementioned system within ABC Insurance Company. The objective of the study is to evaluate the program's influence on sales efficiency and profitability by conducting interviews with a targeted cohort of employees who have employed the software in their sales endeavours. Furthermore, the study will include hypothesis testing to analyse the timing of quotation production across various policy configurations utilising Salesforce Industries CPQ, aiming to provide further validation of its benefits.

In order to better understand the relative costs of IaaS, PaaS, and SaaS, it is recommended that this study collect data on these three categories of cloud computing services. The purpose of this research is to analyze how much money is saved or lost while adopting cloud services for different-sized enterprises. Its purpose is to identify and quantify outcomes that are beneficial to society, whether they be monetary or otherwise. Additionally, we want to provide a complete picture of cloud services' economic effect by itemizing their benefits and drawbacks. The collected data is stored in a secure location and treated ethically to prevent unauthorized access. Then, any conflicts interesting are investigated thoroughly and resolved. The researcher may be certain that the study is being conducted in a responsible and moral manner if they follow these criteria. The research summed up with the conclusion that every device is capable of running any SaaS application. It could be a smartphone, laptop, desktop computer, or tablet. In this manner, businesses can better protect their data, and employees may take advantage of the flexibility of accessing services from any device so long as they have their secure line key.

The Indian pharmaceutical industry has witnessed exponential growth over the past few decades to become the third largest globally in volume terms and a leading supplier of affordable generics. This paper provides a comprehensive assessment of the industry’s evolution, structure, performance metrics, priorities, challenges and future outlook. Secondary data from government reports, industry associations, company documents and academic literature is synthesized to develop a holistic perspective. The analysis indicates that innovation, research productivity, clinical trials, ethical practices and sustainability shape the industry’s development. Strategic priorities include strengthening R&D capabilities, expanding exports, leveraging digital technologies across operations, ensuring robust supply chains and adopting global quality and ESG standards. Investments in innovation have risen but remain below international benchmarks. While generics account for majority revenues, emerging segments like biosimilars and complex drugs offer growth opportunities. Digital adoption is increasing with data analytics and AI gaining prominence. However, issues related to regulations, pricing control, IP protection, skill shortages and environmental impact need resolution through appropriate reforms and public–private collaboration. India’s vision to become a leading pharmaceutical innovation and manufacturing hub would require reinforcing policy frameworks around IP protection, pricing, regulatory efficiency, skill building and technology adoption. Overall, the industry holds immense potential for beneficial growth but realizing it would necessitate concentrated efforts from stakeholders to build organizational capabilities, foster collaboration and imbibe global best practices.

Numerous items are used in daily life, resulting in millions of tonnes of packaging waste being generated daily, including cosmetic containers, cardboard packaging used by various e-commerce services, plastic bottles, carry bags, and so on. With a focus on sustainable development, researchers have been working tirelessly to reduce waste generated from the packaging sector by improving the conventional principle, and this project proposes the 4thR—RETURN, with an aim to develop a promising solution that would work on the fundamentals of the reverse supply chain, that is, it shall collect the reusable packaging from the consumer and deliver it to its manufacturing unit or the industry/warehouse where it can be reutilized. To achieve the project’s goal, various technologies, such as multiple layer perceptron, multiple linear regression, and full stack development are used to build a gateway that can allow consumers to return the packaging in the same manner that many e-commerce websites accept product returns.

This study focuses on deep learning methodologies as it examines the spectrum of sentiment analysis techniques. Identifying the emotional tone or attitude reflected in textual information is called sentiment examination, an important component of natural language processing. A subset of machine learning called deep learning has completely changed the discipline by making it possible to build sophisticated models that can accurately capture complicated linguistic patterns. This study provides a thorough analysis of deep learning methods applied to opinion analysis. In this research, we examine and classify a variety of deep learning methods for sentiment analysis. We also examine current issues and potential developments in sentiment analysis. We also covered assessment metrics, which are frequently used for developing and accessing sentiment analysis models.

Deep learning techniques for classifying images into multiple classes have made significant strides in the past few years. Nevertheless, allocating multiple classes to a single image, as seen in object detection scenarios, remains a relatively unexplored avenue of research. This study is dedicated to harnessing the potential of deep learning methods to discern various diseases within a singular leaf image of a citrus fruit. The investigation focuses on citrus leaves affected by none, one, two or all three severe diseases—anthracnose, melanose and bacterial brown spot. These leaves serve as the training dataset for ten distinct deep learning models. The performance of these models is meticulously examined, gauging their accuracy in classifying the diseases. The outcomes highlight the superiority of the DenseNet121 architecture in terms of accuracy and training duration, as it accurately identifies overlapping disease classes present in citrus leaves. Following suit, the MobileNetV2 architecture showcases comparable accuracy and a noteworthy reduction of 66% in training time.

The health of human is the functional ability to face everyday in life. The human living has changed a lot and in a better way in this modern days. The Internet of Things (IoT) technology melding with healthcare sector affirms every individual good efficiency. IoT in healthcare sector assures a very improved and better treatment as it supports remote monitoring. In remote monitoring, the human body is monitored remotely with less human involvement. This paper includes the work done with MAX30100, AD8232 sensors for heart monitoring. The average error % is 4.04%, and the average accuracy is 96.12% which is the result of analysis.

A trending topic that has grown in prominence over the past several years is the Internet of Things (IoT). With the steadily increasing adoption rate of Internet-enabled devices in applications like smart homes, smart cities, smart grids, and healthcare applications, the demand of IoT becomes increasing. Now this needs to guarantee the safety of data and communications privacy among these IoT devices and their supporting infrastructure. IoT involves numerous low-resource devices, and most of these devices often need to secure interaction with their network administrators, which are IoT network nodes with more resources. Even though more services and applications are being connected through wireless network, security is still a major concern. IoT system security is a subject that is actively being researched. In this research paper, we discussed and contrasted the various network-based IoT authentication and communication security techniques. Few key points are also discussed regarding the IoT challenges and popular IoT cryptography approaches along with lightweight cryptography and trends in IoT.

In today’s internet-driven world, fake news is a problem that is only becoming worse. Given the ease of exchanging information online, separating false information from reliable information is a crucial endeavor. Using bag-of-words and consecutive mining approaches, we provide a data mining solution in this work to categorize articles as genuine or fake. We also compare the accuracy of the solution for identifying fake news across different datasets. Our method first purifies the input information by normalizing words and eliminating “filler” words. The cleansed data is then vectorized using sequential mining techniques. After that, it uses vectorized data to train the classification models and categorizes unknown news as authentic or fake. Assessment of our technology to mine and categorize bogus news using actual data demonstrates its viability. The classification algorithms are then trained using vectorized data to categorize unreported news as real or bogus. The effectiveness of our technology in identifying and categorize bogus news has been evaluated using real-world data.

The concept of reversible data hiding (RDH) enables the full restoration of the cover image while simultaneously recovering the concealed data from a previously obscured image. Hence, it is the favored choice when complete restoration of the cover image is required in situations where the concealment of critical data is important. This work introduces a system that utilizes interpolation-based color coding method (CCM) to temporarily conceal data without permanent deletion. The refinement of the expanded form of the original image is achieved by the utilization of two distinct methods, namely enhanced neighbor mean interpolation (ENMI) and modified neighbor mean interpolation (MNMI). This procedure is conducted prior to the inclusion of any confidential data. The experimental findings indicate that the suggested approach has the potential to be implemented and exhibits superior performance compared to standard methods in relations of highest signal-to-noise ratio and data whacking size.

Image noise is undesirable that can negatively affect the quality of digital images. It reduces the image quality and increases the processing failure ratio. It is highly recommended to remove the noise, and before removing the noise, we have to know the type of noise and estimate the parameters of noise for developing effective noise reduction techniques. This study introduces a method to effectively detect, recognize, and estimate image noise of various types (Gaussian, lognormal, Rayleigh, salt and pepper, and speckle). The proposed model consists of four stages: the first stage is detecting the noise in an image using a convolutional neural network. The second stage classifies the noisy images into one of five types of noise using a new method based on a combination of deep wavelets and support vector machines (SVM) classifier. The third stage involves estimating the parameters of the noise using maximum likelihood estimation (MLE). Finally, choosing the most suitable noise reduction technique for each type using linear and nonlinear filters and showing the capability of the suggested technique in estimating multiple noises commonly present in digital images. The proposed method utilizes a likelihood function derived from the MLE model for each noise type to estimate the noise parameters. Then used to select the most suitable noise reduction technique for each type. The quality of the denoised images is calculated utilizing the peak signal-to-noise ratio (PSNR) as the evaluation metric. The results show that the combination of wavelets with machine learning, specifically SVM, can highly enhance the results, where the accuracy was 93.043% through many experiments conducted to build a sturdy classification model. The MLE-based noise estimation method is also a reliable and accurate method for image noise estimation, especially for Gaussian, salt and pepper, lognormal, and Rayleigh noise. However, for highly noisy types such as speckle noise, further research is required to improve the estimation accuracy. This study contributes to the development of more effective noise estimation methods for improving the quality of digital images.

The energy-efficient path selection algorithm proposed in this paper balances the conflicting goals of maximizing network lifetime and minimizing energy usage routing in mobile ad hoc networks (MANETs). The proposed strategy maximizes lifetime energy efficiency, MANET, and deep learning. Produce the data after building the network by carrying out assaults and validating paths. Then sketch a neural network with capabilities for prediction and performance evaluation. Then nodes in a network that are negative by definition must be followed by choosing the optimum route. Employed in the current study to increase the energy efficiency as well as the kind of data handling on the network with the metrics of stolen time, total time, total energy, and packet delivery rate, predict the energy and lifetime maximization utilizing deep neural networks for deep learning, management, and lifetime energy efficiency maximization. Five hundred packets of data from a neural network were used to get the maximum value. The total energy used is 7570, packets are delivered at 74.60, time taken is 371.81, and the minimum theft rate for 500 packets is 6.8.

This paper presents a system to monitor quality levels as a baseline for assessment by detecting the mixture of gases, vapours and particles in an indoor environment and display them with respect to Air Quality Index (AQI). The present situation in air pollution is mainly due to various physical and chemical activities in industries and vehicles which made the air quality to questionable point. Pollution levels are rising exponentially as a result of factors such as industry, urbanization, population growth and usage of automobiles which all lead to problems harmful to human health. If the air quality drops below a particular threshold, the indicator shows there are harmful gases in the air such as—carbon dioxide, smoke, alcohol, benzene, ammonia and nitrogen oxide. This vital measurement will give appropriate awareness among the public towards healthier life.

Power is one the most fundamental requirements for people in the present. Change of sun-based energy into power further develops age of power as well as decreases contamination because of petroleum derivatives. The result force of sunlight-based charger relies upon sun powered irradiance and heap impedance. Since heap impedance relies upon application, a DC converter is utilized for working on exhibition of sun powered charger. Sun-based irradiance and temperature are dynamic. Thus Internet-based calculation that progressively registers working mark of sunlight powered charger is required. The effective change of sunlight-based energy is conceivable with greatest power point following (MPPT) calculation. The different calculations in MPPT and their geography are talked about in this paper.

Aggressive scaling of MOSFET results in degradation in their performance due to different short channel effects. The problem was addressed by use of high-k dielectric followed by different structural modifications. Among different modified devices, DG-MOSFET seems to be promising one due to scalability, simple structure and offering better control on channel. However, to obtain enhanced performance further structural modification of DG-MOSFET was introduced. This study aims at performance analysis of DG-MOSFET with graded channel and multi-gate. The results clearly indicate reduced DIBL effect for modified DG-MSFET structure.

Optimizing the communication network's performance under diverse service quality constraints to match the briskly expanding claims of wireless/mobile applications is the vital goal of imminent wireless networks. A conspicuous way to improve the network performance is through embodying several scheduling mechanisms. Though various scheduling schemes exist, improved schemes are still needed for performance breakthroughs. Therefore, this article provides intense research on scheduling and performance optimization of communication systems. It outlines the prime scope of scheduling resources and enhancing diverse performance measures for strengthening and facilitating wireless network performance. It investigates the existing studies on resource allotment, scheduling and performance enhancement of communication networks. This review work illuminates some vital performance metrics involved in performance upgradation. The paper finally presents the paramount research challenges explicitly involved in the performance betterment of communication networks for introducing and implementing optimal schemes and encouraging vast research in this direction.

With the help of present state-of-the-art microelectronic and electrical technologies, smart wireless sensor networks (WSNs) can be realized that are capable of sensing and monitoring diverse type of environmental parameters like air temperature, humidity, pollution, gas level, etc. Measured data provides a feedback that is used to operate different support systems like air conditioners and ventilators for environmental control. High energy efficiency along with high network throughput rates and low latency are critical design issues for a wireless sensor network. Different types of network protocols like node geographic location-based, multihop, data aggregation, and hierarchical clustering protocols have been proposed for performance enhancement of WSNs. This paper presents a detailed review of different routing protocols for WSN along with their merits and demerits. Node geographic location-based routing like Geographic Distance Routing (GEDIR) and hierarchical clustering (HC)-based routing protocol such as Distributed Energy-Efficient Clustering (DEEC) provides and increased network lifetime and network throughput performance along with reduced network latency performance of a WSN.

This article seeks to anticipate AirBnB prices using advanced regression approaches. Extensive data analysis was done on different databases spanning diverse variables such as location, property type, facility, and user level. The database is constructed utilizing robust approaches such as feature augmentation, outlier reduction, and value loss. A number of complex regression models, such as linear regression, decision tree, random forest, gradient regression, are generated on the pre-developed database. The model is improved through hyperparameter adjustment to increase prediction accuracy. A cross-validation approach was employed to examine the performance and resilience of the model. In addition, a feature significance study was undertaken to discover the most significant elements impacting Airbnb prices. The experimental findings suggest that the improved regression approach delivers greater prediction accuracy than the standard model. The results of this study add to Airbnb’s pricing system and can promote improved decision-making for hosts and visitors searching for competitive pricing.

It is common knowledge that the Sundarbans tiger reserve, which spans the border between India and Bangladesh, is one of the most important habitats in the world for the endangered Royal Bengal tiger (Panthera tigris tigris). The health of a tropical forest is strongly dependent on the variety of tiger species that live there, as well as their patterns of movement, the richness of their groups across time, and the ecosystem services that they provide. The purpose of this study is to investigate the relationship between tiger migration and the distribution of prey and topological features in this region. We are undertaking a community analysis of the tigers that live in the Sundarbans by looking at the distribution of their prey and mapping the vegetation in the area. We have been provided with datasets that comprise information on the distribution of prey, GPS locations of tigers, and vegetation mapping across the Sundarbans region by the Wildlife Institute of India (WII). The results of the research indicate that prey distribution, the kind of vegetation, and the time of day all play a significant role in determining the movements of tigers as a collective community component, which can further assist researchers to develop measures and strategies for tiger conservation at Sundarbans region.

The financial forecasting of different firms in the area of financial status aims to determine whether the company will go bankrupt in the near future or not. This is a critical problem for these companies. Several companies have shown a strong interest in this area, particularly since they are concerned about the future of their companies from a financial perspective and want to determine whether or not they will go out of business. Therefore, in the work that we have done, we have presented three well-known technologies of deep learning in conjunction with ensemble classifiers and boosting ensemble classifiers for the purpose of failure prediction. During our investigation, we used an uneven dataset consisting of businesses from Spain, Poland, and Taiwan. In addition to this, we applied approaches such as oversampling, hybrid balancing, and clustering-based balancing to get rid of the inconsistent data. When taking into account a real-life financial dataset with an appropriate amount of complexity, it was discovered that the MLP-6L model with the SOMTE-ENN balancing approach had the most remarkable performance when measured against the metrics.

Machine learning approaches came about as a game-changer in modern healthcare, leading to more reliable medical predictions and enhanced patient care. Predicting the way of delivery during labor is critical to protecting mother and newborn health. This study offers a thorough comparison of machine learning (ML) approaches aiming at predicting an optimal mode of delivery. The efficacy of enhanced ML algorithms is evaluated in improving prediction accuracy using a dataset containing maternal health indicators. This study compares the effectiveness of five distinct machine learning approaches: J48, Logistic Model Trees, Random Forests, Random Tree, and Multilayer Perceptron. We analyze their prediction capabilities and applicability for the task at hand through a thorough experimental procedure. Our findings show that different advanced ML techniques have varying degrees’ effectiveness in forecasting the way of delivery. The performance parameters under consideration are accuracy, precision, and recall. Among all the performance metrics considered, J48 exhibited the most favorable performance.

Document similarity refers to an approach of measuring how two or more documents look alike in terms of their content or structure. Document similarity algorithms are used to determine the degree of resemblance or relatedness between various documents. Document similarity plays a pivotal role in a wide range of tasks involving natural language processing, information retrieval, recommender systems and duplicates detection. In this paper, we will be studying and compare the similarity score of documents using different document similarity measures and models like cosine similarity, Euclidean distance, Jaccard similarity, Latent Semantic Analysis (LSA), Latent Dirichlet Allocation (LDA), Bidirectional Encoder Representations from Transformers (BERTs), etc.

With the advent of FinFET-based circuits, the scope of Moore’s law may be extended without the continuous scaling of CMOS devices. FinFET devices provide a viable method to create highly integrated, power-efficient Schmitt Trigger circuit for low-power digital applications due to its combination of enhanced flexibility and decreased short channel effects (SCEs). Schmitt Triggers are used in the design of Integrated Circuits (ICs) to produce digital signals from analogue signals in order to facilitate the implementation of short channel lengths and deliver exceptionally ultra-low power. The Schmitt Trigger Mechanism is being worked on, and the Schmitt Triggers are an important System on Chip (SoC) circuit. The simulations are carried out for various parameters such as leakage, parametric variation of power consumption, and dissipation using Cadence Virtuoso Tools at 45 nm technology. The MTCMOS technique is helpful in reducing the leakage parameters, and it is more appropriate than the conventional one. In the FinFET-based Schmitt Trigger, the leakage power is lowered up to a value of 27%, and the leakage current is reduced by 49% when compared to the conventional Schmitt Trigger.

Computer vision and the area of artificial intelligence (AI) both heavily rely on the detection of facial expressions. This article concentrates on operations based on face images. It demonstrates how visual articulations are most important data facilitates, despite the limitless possibilities of how FER can be analyzed by using various instruments. This essay provides a succinct analysis of recent FER research. However, theoretical FER structure designs and their initial evaluations are displayed close by conventional FER approaches. The presentation of numerous FER views using the “start to finish” learning permission through critical associating authorization follows. As a result, this study will help in connecting a convolutional neural network (CNN) for some LSTM components (long transient memory). This paper concludes with a short poll, evaluation assessment, findings, and standards that serve as a standard for measurable connections between all of these FER studies and experiments. For students in FER, this audit can serve as a succinct manual that provides pertinent details and evaluation for recent tests. Additionally, knowledgeable examiners are searching for promising paths for future work.

A smart city is an innovative, urban, organized, and sustainable city. It mainly depends on the Internet of Things (IoT) technology which improves the excellence of living, safety, the operational efficiency of urban services, decision-making, government services, and social well-being of its people. Smart city means dealing smart in education, government, mobility, households, and e-health. IoT enables all smart devices to connect through the Internet, like sensors, detectors, actuators, wearable’s, mobile phones, watches, and smoke detectors. The rapid increase of Internet of Things in most smart city applications defines new security hazards that threaten the confidentiality and safety of end devices. Therefore, it is significant to improve smart services and the information protection and privacy process. The IoT devices need wireless sensor links and Radio Frequency Identification (RFID) to benefit from IoT. These resource-limited require common authentication between the devices through the association of a novel device, where authentication and encryption of the data to be sent. This paper has three contributions. First, surveying the fundamental smart city privacy problems. Second, the paper proposes a firm, trivial, and energy-efficient security solution for IoT systems which is called; New Compact-Data Encryption Standard (NC-DES). Finally, a case study of the healthcare framework is taken as an example of IoT applications to provide secure transfer of the measured parameters. The proposed system proved that the IoT devices had been secured without wasting their limited resources.

The present achievement of Unmanned Aircraft System [UAS] innovation has led to it being the answer to every issue that arises in daily life. Hexacopter technology is now expanding quickly across a wide range of application areas. Development, design, and evaluation of hexacopters are carried out in the work using previously acquired actual elements. Drone autonomy, beyond visual line of sight control, and practical approaches for hexacopter platforms are all covered in depth, as are the mathematical modelling, controller, and command and control system design for multicopter drones in general, with an emphasis on hexacopter systems. The mechanical foundation of a physically constructed hexacopter, together with elements of matrix-based mathematical modelling principle, is presented, along with a video of the drone in flight.

A decade of research has been conducted on detecting copy-move forgeries (CMFD). Technology has enabled the manipulation of images, once the most authentic source of information. This paper proposes a copy-move forgery detection algorithm based on fused features to address issues such as time complexity and difficulty detecting forgeries in smooth regions. To extract descriptive features, a low-contrast threshold was used in conjunction with three detection methods, including scale-invariant feature transform (SIF), speeded-up robust features (SURF), and accelerated KAZE (AKAZE). SURF and accelerated KAZE (AKAZE) are used in our keypoint-based CMFD technique. To detect manipulated regions efficiently, AKZAE, SURF, and SIFT can be used to extract major keypoints in smooth regions.