Search

Natural disasters have significantly increased in occurrence and severity, causing substantial destruction and loss of human lives. Existing research predominantly focuses on the immediate impacts of disasters, lacking in-depth analyses of long-term financial market repercussions. This study aims to fill this gap by examining the effects of natural disasters on the stock markets, specifically focusing on hydrology, geophysics, meteorology, and climatology disasters. This study reviews the literature on the effects of natural disasters on the financial market by utilising the Scopus database from 2018 to 2023 and focuses the discussion on three areas, i.e., the relationship between natural disaster events and the stock market, methodologies, and the geographical area of the study. The findings indicate that stock market response varies based on criteria such as the type of natural hazards, geographical location, and industrial sector. Regression analysis, GARCH models, and event studies have emerged as the prevailing techniques employed in this field of study. The findings indicated that most studies concentrated on developed economies such as the US, Europe, and Japan. Future research should consider a broader scope of topics, including emerging nations, and utilise more sophisticated analytical methods. The findings of this study are essential for improved decision-making and risk mitigation strategies in the face of escalating climate change risks. [ABSTRACT FROM AUTHOR]

The study's overarching goal is to evaluate the efficacy of two different algorithms for SAR images categorization: the K-nearest neighbour technique and a new convolution neural network. We extracted 20 data instances from the Kaggle database system. The dataset is separated into two sets as training and testing. In 20 dataset 10 used for training and 10 used for testing. Classification of synthetic aperture radar pictures and comparison of K-nearest neighbours' and the unique convolution neural network's identification efficiency are conducted in MATLAB. The sampling number is determined using the G power formula, with the variables alpha set at 0.05 as well as power at 80%. From the MATLAB simulation the novel convolutional neural network algorithm obtained a classification accuracy of 95.31% and K-nearest neighbour algorithm obtained classification accuracy of 73.01%. Utilizing SPSS evaluation, we found a significance value of 0.005 (p<0.05). In conclusion, the new convolution neural networks outperform the K-nearest neighbour method in terms of effectiveness on the given dataset for artificial aperture radar picture classification. [ABSTRACT FROM AUTHOR]

This study set out to evaluate the state of the drought in the Kancheepuram district by means of the VCI and NDVI indexes. Data samples were obtained from the MODIS NDVI database. Group 1 consisted of twenty samples, while Group 2 also consisted of twenty samples. The computations were carried out assuming a beta of 0.2 and a significance level of 0.05, with a 95% confidence interval and G-power 0.8. The water sample did not pose any health risks to humans. Twenty people evaluate the situation using two drought variables: the Vegetation Condition Index (VCI) and the Normalized Difference Vegetation Index (NDVI). For instance, the seasonal rainfall index may display an NDVI of 0.40 along with a VCI of 0.70. This study is considered statistically significant since the p-value is less than 0.05 (or p=0.021). When assessing the drought situation in the Kancheepuram district, VCI yields more accurate results in the seasonal variation index than the NDVI index. [ABSTRACT FROM AUTHOR]

By contrasting Sobel edge detection with Prewitt edge detection, this study hopes to enhance the precision of dynamic data for identifying forest fire boundaries and flames. The data used in this study is taken from the freely available Kaggle database. A more accurate forest fire detection and monitoring system's sample size was determined using G-power 0.8, alpha=0.05, beta=0.2, and a 95% confidence interval. There were ten participants in Group 1, and ten in Group 2. Our improved forest fire monitoring and warning system makes use of both Sobel and Prewitt edge detection, both of which use the same number of samples (N=10). The accuracy rate of the Sobel edge detection classifier is 85.446% greater than that of the Prewitt classifier, which is 76.254 percent. A significant threshold of (p<0.05) is used in the research, which is equivalent to p=0.022. When it comes to spotting forest fires, Sobel edge detection has a higher accuracy rate than Prewitt edge detection. [ABSTRACT FROM AUTHOR]

In the end, we want to use a database of chest X-rays together with some deep learning-based machine learning to discover novel COVID-19 predictions. With the use of computer-aided detection, chest X-rays may reach this level of precision. Nevertheless, there has been little focus on the latest CAD results for detecting new corona virus predictions and concurrent pneumonia on CXRs. The Kaggle library accurately detects the deletion of COVID-19. We chose 30 examples out of a collection of 2,000 chest X-rays. Alexnet and Googlenet are the names of the two groupings. According to the results, Alex net outperformed Google net with an accuracy of 96.18%. Alex net performs better than Google net over the long term. Two approaches used in the independent sample t-test meet the statistical criteria with a 90% confidence level and a p-value of 0.001 (p<0.05). According to the results of the research, the main goal of the project is to develop and improve algorithms for predicting the spread of COVID-19. Alexnet outperforms Google net, according to the results. [ABSTRACT FROM AUTHOR]

Examining how well two edge detection algorithms, namely Prewitt edge detection and the novel Sobel approach, can use dynamic and chromatic data to spot forest fires and flames is the main objective of this study. Materials and Methods: The open-source Kaggle database is where the data used in this study is retrieved. We took ten samples from Group 1 and ten samples from Group 2 to assess the forest fire detection and monitoring system's enhanced accuracy rate. Twenty samples were used in all. We used G-power 0.8 with a 95% confidence interval, and we set the alpha quality at 0.05 and the beta quality at 0.2. To improve the accuracy of forest fire identification and tracking, a 10-item sample was used to combine the new Sobel edge detection with the tried-and-true Prewitt edge detection. The results show that the new Sobel edge detection classifier achieves a higher accuracy rate than Prewitt's 78.135 percent. P = 0.035 is equivalent to the study's predefined significant criteria of (p < 0.05). Sobel edge detection is more effective than Prewitt edge detection in identifying forest fires, according to the results. [ABSTRACT FROM AUTHOR]

This study primarily aims to compare the performance of Convolutional Neural Networks (CNNs) and Decision Trees (D-Trees) in detecting cancer cells in x-ray images of the bone. This study makes use of data that is already accessible to the public in the NTHU Computer Vision Lab database. Both Group 1 and Group 2 had 140 photos. The computation was performed using G-power 0.8 and used a 95% confidence interval, an alpha of 0.05, and a beta of 0.2. We used 280 photos as our sample size for tumor cell identification and classification utilizing bone x-ray photographs. Convolutional Neural Networks (CNNs) and Decision Trees (D-Trees) categorize cancer cells in bone x-ray pictures with a sample size of 10. Compared to the Decision Tree (D-Tree) classifier, which had an accuracy rate of 91.0934%, the CNN classifier achieved a superior rate of 97.9034%. A p-value of 0.022 indicates that the research was statistically significant. Convolutional Neural Networks (CNNs) outperform Decision Trees (D-Trees) in effectively detecting and classifying malignancy cells from bone x-ray images. [ABSTRACT FROM AUTHOR]

The use of artificial intelligence methods in the classroom has become increasingly common. The advancements in artificial intelligence have had far-reaching consequences and have pushed for new approaches to education. The full integration of different artificial intelligence techniques into education has shown that distance education and learning are open and flexible. The application of artificial intelligence in the field of education has emerged as a prominent topic of study in recent years, making an exploratory review of the subject matter necessary. This study examines data extracted from Scopus database which yields 918 research articles on artificial intelligence in distance education and distance learning published up to 2022. The aim of this study is to identify trends and themes of the research area. The frequency of words used in the titles and abstracts of the articles has been identified using bibliometric analysis, which has also been used to determine publication trends. Yearly trends of publication and most cited articles during the publication period have been identified. This study used the topic modeling method to find coherent research topics that are the focus of the articles and were used to find research themes. There are ten distinct themes that have been uncovered using the Latent Dirichlet Allocation technique. This study makes a contribution toward the goal of providing a clear view of the existing themes or topics of research related to Artificial Intelligence in the field of Distance Education and Distance Learning. [ABSTRACT FROM AUTHOR]

This bibliometric study aims to provide an overview of the global research trends in the field of cultural heritage from 2013 to 2022. The dataset used for this study was obtained from the Scopus database, which includes a vast collection of scholarly literature from various disciplines worldwide. A total of 5,901 research articles were identified, and an analysis was conducted to evaluate the productivity, impact, and trends in cultural heritage research. The study found that the annual growth rate of scientific production has consistently increased since 2016, reaching its peak in 2022. Italy emerged as the most productive country in this field, and the Consiglio Nazionale delle Ricerche (CNR) was identified as the most productive institution. The study also reveals that Tsvi Kuflik, affiliated with the University of Haifa in Israel, was the most prolific author, having authored 24 documents on the topic. Additionally, co-occurrence network analysis using keyword plus was used to visualize the relationships between different research topics in cultural heritage. The visualizations indicate a bifurcation of the network into two primary clusters, with blue and red representing "cultural heritage" and "cultural heritages," respectively. The study highlights the importance of cultural heritage research in promoting social cohesion, preserving cultural diversity, generating revenue and job opportunities, addressing social and environmental issues, and providing insights into human history. The findings of this study contribute to the understanding of the research landscape in cultural heritage and can assist researchers in identifying critical areas of focus and potential interdisciplinary linkages. [ABSTRACT FROM AUTHOR]

Over the past decade, researchers have noticed immense interest in cryptocurrencies. The effects of cryptocurrency-based research on the Scopus database must be examined in this area. This study uses a method that combines bibliometrics and content analysis to find the countries, publications, authors, and articles that talk about cryptocurrency and how it might be used between 2013 and 2022. A totally of 1370 different articles were then included in the bibliometric study. The literature on cryptocurrencies is constantly evolving. Development continued for the first five years, 2013–2017, with fewer than 15 articles per year. However, starting in 2018, it skyrocketed, peaking in 2022 with a total of 458 articles in Scopus. The US was the most prolific country in cryptocurrency research during this period. With six papers, Giulia C. Fanti from CMU in Philadelphia, USA, became a prolific author. The Ho Chi Minh City University of Economics in Vietnam has written 21 articles about cryptocurrency and has become a well-known research partner. Elie Bouri from Lebanese American University, Beirut, Lebanon, is the most prolific author in cryptocurrency research from 2013-2022, with 15 articles in the Scopus database. Not only is he prolific, but he is also the most influential author, with 823 citations in four years. Until 2020, bitcoin-related research dominated, followed by blockchain research. Nevertheless, in 2021, cryptocurrency research related to blockchain will dominate. This indicates that research related to cryptocurrency is evolving toward its security. [ABSTRACT FROM AUTHOR]

The main goal of this study is to conduct a thorough bibliometric analysis of the e-learning domain, along with a descriptive analysis. This study has a specific focus on publications that were included in the esteemed Scopus database during the pandemic period spanning from 2020 to 2022. The study entailed the classification of scholarly articles on electronic learning by means of bibliographic data sourced from the Scopus database. A comprehensive search was conducted on the Scopus database using the keyword "e-learning" with a title restriction, resulting in a total of 3,556 documents for the time frame spanning from 2020 to 2022. Descriptive and bibliometric analyses were conducted utilizing the Biblioshiny tool based on the data that was gathered. The findings of the analysis indicate that the majority of the publications were disseminated in 2021. Indonesia holds the largest proportion of publications, trailed by India and Malaysia, thus positioning it as the country with the highest quantity. The topic has been extensively researched and documented by Ragad Moufaq Tawafak, Sohail Iqbal Malik, and Ghaliya Musalam Salim Al-Farsi, who are esteemed scholars affiliated with Al Buraimi University College located in Oman. The literature on e-learning indicates that it is a multifaceted phenomenon with a range of factors contributing to its success or failure, as evidenced by the most commonly cited publications on the topic. [ABSTRACT FROM AUTHOR]

This study aims to assess legal research in Indonesia by analyzing the legal research written by Indonesian authors from 2002 to 2022. The study uses a bibliometric method and the Scopus database to identify 616 documents in the form of journal articles and conference proceedings. The results show that legal research in Indonesia is still in the process of modernization, and there is a need for more consistency and coherence in the laws and regulations. The study also found that the annual growth rate of legal research in Indonesia is significant at 36.99%. Moreover, it revealed that the most common keywords used by authors are "legal protection" and "Indonesia," and Islamic law and legal certainty issues are trending research topics. The study concludes that studying legal research in Indonesia is beneficial for anyone interested in understanding and applying Indonesian legal principles and for those who want to develop skills that will be useful in many fields. [ABSTRACT FROM AUTHOR]

The world has entered the era of industrial revolution 4.0. The Internet of Things continues to be developed to support the industrial revolution and the utilization of more appropriate technology to help human work. IoT-based baby incubator temperature and humidity monitoring system with the ultimate goal of producing a research that in the future can be developed as a viable system for medical purposes. The system is designed as a prototype and props by utilizing the DHT11 sensor as well as relay output modules, buzzers, and lights. Raspberry Pi acts as a data processor that receives data from the sensor and then stores it in the database that has been embedded in it. The database can be displayed in real-time on a web page designed for all devices. The lamp as a temperature guard props is automatically controlled and the buzzer module as an alarm if the system is not connected to the network. The final goal of this research is to produce a simple monitoring product that can record baby incubator data and the data can be developed as an analysis by medical staff later by utilizing IoT and machine learning. [ABSTRACT FROM AUTHOR]

A website is a page containing information that can be viewed if the computer is connected to the internet. With a website, everyone in the world can get and manage information from various sources available on the internet. The website itself can currently contain various kinds of media, ranging from text, images, sound, and even video. With the existence of a website via the internet, Mandiri Arts Crafts decided to use the website as one marketing medium so these products and products can be wider and known to the wider community. The people around and especially the city of Medan also don't have to bother getting the desired product and don't need to come directly to the product distribution point which is very attractive, beautiful, and worth selling. Study This paper discusses an online-based marketing information system for handicrafts from water hyacinth at Mandiri Arts Crafts. This system helps Mandiri Crafts to update the marketing system for handicrafts from water hyacinth, which is still done manually, so it is inefficient and far from the reach of the public, and its products are still limited to being online-based. This system consists of a system for inputting handicrafts, a system for marketing handicrafts, and preparing reports. The results of this study indicate that the design of an online-based handicraft product marketing information system can help Mandiri Seni Crafts in marketing its products outside the region and facilitate the delivery of information to customers. The study this using Hypertext Preprocessor or only PHP is a language script with function especially general _ used for web development. As well as using MySQL, which is a database management system that is useful for managing databases on websites that will later become data storage areas [ABSTRACT FROM AUTHOR]

This study aims to develop a web-based fish dryer information system to assist and facilitate fishermen in facilitating the drying process of dried fish. The fish drying process here uses a tool developed by applying the Internet of Things concept. The information system will be integrated with the fish dryer so that it is possible to monitor the drying process. The data exchange communication between the dryer and the system is used by the web API service and uses the JavaScript Object Notation (JSON) format. The method used for software development is the prototyping model. The system is built using the Laravel framework, and the data storage process uses the MongoDB database. The information system built has three user levels: admin, fisherman, and fish dryer. The information system will record the data sent from the fish dryer and process the data to display the information needed by fishermen. The implemented information management includes the information dashboard, fish dryer machine data, fishermen data, and historical data. Testing on the system using the black box method. The test has ten scenarios, showing the system can run as expected. [ABSTRACT FROM AUTHOR]

The rapid development of artificial intelligence in predicting the three-dimensional structure of proteins has broken down several important problems that have been faced by experimental structural biologics for decades. Structural biology has assisted in the discovery and development of many new drugs based on structural design. X-ray crystallography and Cryo-EM which have been the gold standard for determining the structure have many challenges in the cost of obtaining large quantities of pure protein samples, the emergence of AI has helped these problems. However, the current artificial intelligence technology actually still requires a database submitted by an experimental biologist for further machine learning material. In the future, stronger collaboration between AI and experimental structural biology is needed in the development of AI for structural biology. This collaboration bridge will provide many advances in the structure-based drug design and become an interesting technique that needs to be understood for integrative structural biology. [ABSTRACT FROM AUTHOR]

The Extraction of data from videos is crucial and likely to be used in many fields for investigation and data collection purposes. Video analysis has been one of the topics for debate for a long time as much data can be extracted from a video. In this article, we have made a prototype of video analyzing software that extracts useful data from the video with the help of various techniques. Our project primarily focuses on the analysis of YouTube videos. To create a user- friendly graphical interface for our tool, we have utilized the Python Tkinter library. We also used the Matplotlib library for data visualization. Deep learning algorithms using TensorFlow, scikit-learn, pandas, CV tool, and imutils libraries for object and action analysis are utilized in our tools for data extraction from the videos. Our tool incorporates sentiment analysis as well, enabling us to analyze the emotional tone of a video and measure its impact on viewers. The data extracted from the video is stored in a database for further processing. This tool can be developed to be utilized in real-time video analysis. Overall, our goal with this project is to provide an automated video analysis for cyber-security by leveraging these powerful algorithms and technologies. [ABSTRACT FROM AUTHOR]

The objective of this study is to improve the precision and energy efficiency of Cognitive Radio Network (CRN) technology design by comparing a new energy detection technique with the traditional Waveform -Based Detection Technique. The experiment entails implementing the created model, which enhances the efficiency of CRN system design through the innovative energy detection technique. The proposed model is implemented using Python, and it utilises a network dataset obtained from the UCI Machine Learning Repository database. The accuracy of Energy Detection is evaluated using a statistical power of 0.8, significance levels (alpha and beta) of 0.05 and 0.2 respectively, and a confidence level of 95%. The calculation necessitates a sample size of 140 individuals, evenly divided with 70 individuals in Group 1 and 70 individuals in Group 2. Within the domain of cognitive radio networks, the Energy Detection Technique (EDT) is expected to surpass the Waveform Detection Technique in accurately estimating spectrum sharing. The proposed CRN system design is assessed based on two metrics: accuracy rate and energy efficiency (EE). The Energy Detection Technique surpasses the accuracy requirement, achieving a confirmation rate of 95.2713 percent. Moreover, the statistical analysis reveals that there is a statistically significant difference of 0.561 between accuracy and loss (p>0.05). The results of the Independent Sample T-Test indicate that the two algorithms are statistically indistinguishable. Overall, this study's analysis of the network dataset demonstrates that the Energy Detection Technique model surpasses the Waveform -Based Detection Technique in terms of energy efficiency. [ABSTRACT FROM AUTHOR]

One of the most prevalent autoimmune diseases that can affect the Central Nervous System (CNS) in mild to severe ways is multiple sclerosis (MS). The goal of this study is to separate the MS lesion from two-dimensional in nature (2D) brain MRI sliced of each modality with different planes (axial, coronal, and sagittal) using a preconditioned ResUNet segmentation technique. This work aims to provide a better MS detection by implementing CLAHE supported image enhancement scheme to improve the visibility of MS section in the chosen image. A trial of this scheme is conducted on 25 patient's database (500 images with dimensions of 512 × 512 × 3 pixels) and the achieved performance values are considered to verify the merit of the proposed scheme. The achieved result of ResUNet scheme is then compared and confirmed with other methods, such as Unet, SegNet, VGG-Unet and VGG-SegNet found in the literature. The experimental outcome of ResUNet with ResNet18 as the backbone helps to get a better value of Jaccard (>86%), Dice (>93%) and segmentation accuracy (>98%) on the chosen 2D MRI slice. The proposed work is simple and in future it can be considered to examine the clinical grade MS database. [ABSTRACT FROM AUTHOR]

The purpose of this study is to use new Artificial Bee Optimization methods over Particle Swarm Optimization (PSO) techniques to increase the Energy Efficient Wireless Sensor Networks. The WSN-CENS provided the research dataset that was used in this study. WSN-CENS (Wireless Sensor Networks - Center for Embedded Networked Sensing) database system with 126 samples. In which 75 % of data is dedicated to training, and 25 % is for testing. The energy efficient clustering is to be assessed with an iteration size of N=10 (Both Group 1 and 2), and the computation was carried out using a G-power of 0.8, alpha and beta values of 0.05 and 0.2, and a 95% confidence range. It was discovered that the Particle Swarm Optimization method had a success rate of 92.75%, while the Novel Artificial Bee Colony approach had a better accuracy rate of 94.35%. According to the statistical study, there was a p=0.01 difference (p<0.05) between the two methods. The accuracy of the novel Artificial Bee Colony algorithm is noticeably higher than that of the Particle Swarm Optimization method. [ABSTRACT FROM AUTHOR]

This article's main objectives are to recognize the sequence in the image, distinguish the object it is, and analyse it correctly. Using the nearest neighbor classifier and the novel random forest classifier, the input picture is used to predict the image recognition. The Kaggle database served as the source of the study dataset for this examination. larger accuracy was predicted for visual pattern analysis (with a sample size of 10 from G1 and 10 from G2) with a sample size of 20. The computation involved the use of a 95% poise interval, an alpha and beta value of 0.2 and 0.05, and a G-power of 0.8. With 91.54 percent exactness, the suggested novel RF outperforms the latter, which has an exactness pace of 85.33 percent. p = 0.001 (Independent Sample T Test p = 0.05) indicates the statistical significance of the difference between the two algorithms. Data analysis shows that for image pattern recognition, the novel random forest model that has been proposed performs better than the K nearest neighbor algorithm. [ABSTRACT FROM AUTHOR]

In contrast to the Linear Support Vector Machine (LSVM) algorithm, the objective about the investigation is to increase the accuracy rate of automatic speaker detection by applying a unique MLP Multilayer Perceptron algorithm. The National Centre for Human Language Technology (NCHLT) database provided the research dataset for this study, which had a G-power of 0.8, alpha and beta values of 0.05 and 0.2, and a 95% confidence interval. Automatic speaker recognition is carried out using MLP and SVM, both of which have the same number of data samples (N=10). MLP, however, has a higher accuracy rate. In comparison to the system, the suggested Multilayer Perceptron gets a success rate of 95.66 percent. higher degree of precision compared to the Linear Support Vector Machine (LSVM). Based on a sample size of twenty (ten from Group 1 and ten from Group 2), the enhanced accuracy rate of automatic speaker recognition was calculated, and the result was an 88.90 percent success rate for the linear SVM classifier. This indicates a noteworthy distinction. The study's significance level was found to be p=0.029. In comparison to Linear Support Vector Machine (LSVM), the proposed Multilayer Perceptron (MLP) model offers a higher level of accuracy for the performance analysis of automatic speaker recognition. [ABSTRACT FROM AUTHOR]

The goal of this research project is to replace CD approaches with an energy detection methodology, therefore increasing the efficiency of energy and precision of the CRN systems design. To improve the efficiency of the CRN system design, a comparison between the newly developed model and the CD approach is made. The proposed model was processed using Python programming, and network data was utilized for the study phase. This investigation was based on online data models from the University of California, Irvine's machine learning database. The computation was run utilizing a G-power of 0.8, alpha and beta values of 0.05 and 0.2, and a 95% confidence interval. 140 people were included in the prediction of the ED precision rate; 70 participants belonged to Group 1 and the remaining 70 to Group 2. The new EDT and the CDT were the two methods utilized to estimate the spectrum allocation for cognitive radio networks; the CDT showed a higher precision rate. The rate of precision and energy efficiency were utilized to assess the CRN System Design's performance. The statistical analysis revealed that the difference between the two methods was not of statistical significance, with an insignificant difference of 0.765 (p>0.05) for both accuracy and loss. The recommended ED Method achieved a 95.2713% accuracy rate. [ABSTRACT FROM AUTHOR]

To use the Trickle Approach rather than the Lift and Shift approach in order to address the low efficiency issue that arose during the database migration and guarantee that the entire database is successfully moved in a shorter length of time. For in-depth testing, the Lift and Shift and Trickle Migration approaches are employed. This experiment is conducted using a sample size of 10. The IBM Employee Dataset, which has 1,059 rows of data, is used as an example dataset. In this experimental investigation, the trickle migration technique outperforms the lift and shift approach in terms of cost, time, and data quality. With an efficiency of 77.63% for the Lift and Shift strategy and 86.20% for the Trickle migration approach, the experiment shows the efficacy of these strategies. According to the data reported in this research, the Lift and Shift strategy is less efficient than the Trickle Migration approach. [ABSTRACT FROM AUTHOR]

Aim: The goal of the proposed study is to compare the overall performance of chosen classifiers and use two distinct machine learning (ML) algorithms to conduct hand gesture detection in real-time to assist the deaf. For this study, neural networks namely K-means clustering, and Support vector mechanism are chosen. Materials and methods: The choice of the data collection, training, and testing of the chosen data using optional classifiers are all part of this phase. A dataset volume of 80% is used to train the suggested ML classifier model, while 20% is used for testing. The output of binary classifiers is divided into two sets for SPSS analysis, with each set consisting of 10 output values under various practical operations, for a total of 20. Next the metrics CI and alpha are valued at 0.292 and 0.95, p<0.05 respectively, during the SPSS analysis. The dataset for the present work contains 1000 unique values related to the light condition and hand position and the factors considered are reviews. Result: The accuracy improvement between both classifiers is compared throughout the experimental investigation using the Python compiler. The accuracy of the chosen Novel K- means clustering, which increased recognition of hand gesture signs, was 96.01%, compared to the Support Vector Mechanism classifier's accuracy of 90.33%. It demonstrates that there is a statistical implication difference amid the K Means Clustering and SVM algorithm with p=0.000 (p<0.05). Conclusion: The hand gestures in complicated backgrounds from the UCI database are used to test and validate an unique hand gesture identification system that has been developed in this research effort. Skin color detection was being employed to segment the hand motion, and morphological operation has subsequently been done to exclude any additional potential skin color detection regions. The system was put to the test using a wide variety of photos that were scaled and rotated, and it performed with an overall accuracy of 96.01%. [ABSTRACT FROM AUTHOR]

The work seeks to intensification the accurateness of identifying human emotion from audio data through comparing a innovative Artificial Bee Colony (ABC) technique toward Particle Swarm Optimization (PSO) algorithm. Dataset for this study was provided by the Toronto Emotional Speech Set (TESS) database, and we employed a example size of twenty (ten from Collection 1 and ten from Collection 2), computed using G - power of 0.8, alpha and beta values of 0.05 and 0.2, respectively. Group 1 uses the Artificial Bee Colony (ABC), whereas Group 2 uses Particle Swarm Optimization. Both groups employ identical number of data samples (N = 10) to predict human reaction identification after audio signals. ABC demonstrated a well level of exactness. The recommended artificial bee colony has an achievement rate of 96.80 percent, which is significantly higher than the particle swarm optimization classifier's rate of 92.25 percent. The learning yielded a p-value of 0.001 (p-value < 0.05), suggesting statistical significance between groups. The proposed Artificial Bee Colony model outperforms Particle Swarm Optimization in the efficacy evaluation of human emotion identification using audio data. [ABSTRACT FROM AUTHOR]

The automatic flaw identification of steel structures in (SHM)structural health monitoring is still difficult due to the presence of a complex backdrop, non-uniform lighting, uneven forms, and interfering in pictures. Traditional methods of defect identification rely heavily on human inspection, which is both labour-intensive and prone to making mistakes. Delays in the removal of deflections as well as the scheduling of the manufacturing process may create an excessive amount of goods to have deformations, which has a negative impact on the quality of the products. A machine learning models is constructed using transfer learning strategies and acoustic emission data was developed to identify fractures in ceramic plates. The ceramics experimental plates were hit with an impact pendulum of the same magnitudes in order to acquire noise curves, which were then used to compile a data set. Following that, the characteristics of this weld defects database were retrieved using the VGG16 and ResNet50 CNNs, each of which had prior training on the ImageNet database. The collected characteristics were then used to train machine learning models including regression analysis, (SVM)support vector machine, and random forest. These characteristics are merged together to create new fusion feature vectors, which are then used to categorise the various weld faults. According to the findings, the performance of the deep features that is based on stacking auto encoder networks is superior to that of the other features. When comparing the two distinct filters, there was no discernible difference in the accuracy of the categorization. Furthermore, the processing learning time required is much longer for lower filter sizes as compared to those for higher filter sizes. In general, the suggested technique has been effective in attaining an accuracy rate of 94.5 percent when identifying the various kinds of cracks. [ABSTRACT FROM AUTHOR]

The rise in home burglary highlights the need for robust security systems. Traditional security systems often rely solely on perimeter protection, which may not be enough to deter intruders who have already gained entry. This research proposes an innovative anti-theft flooring system to enhance home security. The system activates when the homeowner is away, providing an additional layer of protection beyond traditional perimeter alarms. The system leverages real-time monitoring and prompt alerts to ensure quick responses to potential breaches. It utilizes a pressure-sensitive floor mat as the primary trigger, activating the system when unauthorized weight is applied. Additionally, a real-time face recognition system, developed using Python and OpenCV, captures images and compares them to a pre-defined database of authorized faces. This multi-layered approach aims to identify potential intruders in real time, increasing the precision and effectiveness of security measures. A comprehensive evaluation process involving various sensors like pressure mats, motion detectors, and buzzers helps assess system performance. The evaluation not only identifies areas for improvement but also strengthens the overall system by confirming its efficacy in detecting unauthorized access and triggering timely alerts for homeowners. This research contributes to the development of dependable IoT-based home security systems by introducing a multi-sensor approach that goes beyond traditional perimeter protection. [ABSTRACT FROM AUTHOR]

It has been observed that disabled people like deaf and mute have hard time communicating and coping up with normal people due to their inability to speak. So, the only way in which they express their emotions or convey their thoughts is by using sign language, but in today's world very few people understand sign language. Approximately 10% of the world's population is mute or deaf, hence they use sign language as their primary language. Looking at their struggle to communicate led us to the idea of building this project. In this era of advancement, it is very essential to use our technology wisely and in a feasible way, such that it can help people of various stands in their daily life. This project involves a smart glove embedded with multiple flex sensors for precise figure bending or movement of fingers which will send data to microcontroller ARDUINO. The data thus obtained gets processed in microcontroller. Based on that, the hand sign language will be decoded using a predefined database. After this the output gets displayed in an on-board LCD display for the user to view and also gets transmitted in audio and text format to smartphone of the person user's talking to. [ABSTRACT FROM AUTHOR]

The purpose of this study is to examine the relationship between the return of the SET index and the return of the SET index in each industry of the companies that are grouped into eight industries and the return of searches for COVID-19 words in Thailand retrieved from the Google Trends database, which was used as an exogenous variable. Our focus has been on four periods: the first started on 2 January 2020 and ended on 1 June 2020, the second started on 2 November 2020 and ended on 1 March 2021, the third started on 1 March 2021 and ended on 15 June 2021, and the last started on 16 June 2021 and ended on 1 November 2021. The study applied Auto Regressive Integrated Moving Average (ARIMA) with the exogenous variable model (ARIMAX) as the model for measuring their relationship. According to the results, Google Trends for COVID-19 searches in Thailand affected the SET index in each industry in the first period and the second period. According to the results, Google Trends for COVID-19 searches in Thailand affected the SET index in each industry in the first and third periods. In the first period, Google Trends for COVID-19 searches in Thailand affected six industries which include Agro and Food, Consumer Products, Financials, Property and Construction, Services, and the last one is Technology. While in the third period we found Google Trends for COVID-19 searches in Thailand affected six industries including Agro and Food Industry, Consumer Products, Financials, Property and Construction, Resources, and Services. Google Trends searches for COVID-19 words in Thailand during the second and fourth periods did not show any significant variables. Google Trends searches for COVID-19 words in Thailand. The financial industry had a tremendous decrease and was more sensitive than other industries. [ABSTRACT FROM AUTHOR]

A sustainable strategic plan in higher education institutions (HEIs) requires inputs and facts that can be obtained from analyzing the big data. HEIs collect numerous data from students every enrollment, creating big data. However, big data should be given more attention. This study aimed to explore the big data of the postgraduate students in a Malaysian University, namely University Teknologi MARA (UiTM), analyze data characteristics, patterns, and anomalies, and forecast future enrollment models using the Exploratory Data Analysis (EDA) technique. The data were extracted from the university academic database, Student Information Management Systems (SIMS). Several parameters were set in the Structured Query Language (SQL) to mine the data and exported to the Excel software for data post-processing and analysis. The single parameter fitted polynomial regression method was used to develop a forecasting model for postgraduate enrolment from 2022 to 2030. In this paper, we presented the data characteristics and pattern of the postgraduate student demographic and discussed the summary of the postgraduate students' achievements in depth. The developed enrolment forecast modeling shows that the university will have approximately 50,000 postgraduate students enrolment in 2030. Thus, the university must have a strategic plan to identify the future research area, acquire suitable expertise, and upgrade the supporting facilities to accommodate future needs. [ABSTRACT FROM AUTHOR]

The pipeline network constitutes the most important element in the transportation and distribution of petroleum products, so it is important to take care of their safety and check them periodically. Assessment of various faults in pipes. The number of waveform inspection orders in manufacturing, and in-service inspection is growing rapidly. The paper will provide an overview of the basic ideas and various applications that are currently under study and also constitute a database for the reader about the technology to enable him to understand the method and principle of the technology's work. Software that will make targeted ultrasound screening widespread in the 21st century. In addition, the review will provide a brief overview of the history of guided wave and basic ultrasound physics, including modes of wave propagation, piezoelectric action, magnetic stenosis, an electromagnetic transducer, and the general theory of elasticity. All of the above will provide the reader with an important overview and summary to understand the complete research that presents the development of a ring transducer working with Electromagnetic Acoustic technology, which will provide complete coverage of the tube is simple and available and easy-to-use tools to reduce the total cost of the examination. [ABSTRACT FROM AUTHOR]

Numerous artificial intelligence techniques can be utilized in photogrammetry applications. In recent years, the cooperation between artificial intelligence techniques and photogrammetry applications has expanded. This article discusses photogrammetry applications based on artificial intelligence techniques, such as 3D Modeling, Classification, and Camera Orientation. and discusses the outcome of the most effective approach. Automate photogrammetric applications including camera orientation and calibration, classification, categorization and 3D Modeling, information collection for database, and analytical activities. Applications in photogrammetry can be automated with artificial intelligence to reduce effort, save time and cost, and improve work efficiency by increasing accuracy and speed in completing tasks. Finally, we discussed our findings concerning future work on fully automated. [ABSTRACT FROM AUTHOR]