Journal Of AL-Turath University College

Eulerian Video Magnification (EVM) is a technique that enhances and enlarges fine details in videos. It analyzes the temporal signal of the video, magnifying small pixel-level vibrations. By employing image processing and frequency analysis techniques, EVM improves quality and enlarges fine details. Some studies have combined EVM with virtual microphone technology to analyze micro-vibrations caused by recorded sound. These vibrations are used for sound restoration, where they are converted into an audio signal using signal processing and frequency analysis techniques. The resulting sound is then restored in the video. In this work, a literature reviews have demonstrated promising results in terms of accuracy and quality in sound restoration.

digital technology is the main factor shaping the features of this era and her entry into the political field is considered the most influential on the life of society. As it entered the electoral process in all its stages to ensure efficiency, accuracy and speed in its completion. Free and fair elections are the basic foundation for building a democratic system and the legitimacy of the authority in any country. Therefore, this research came to demonstrate the extent to which digital technology can be used in the electoral process Iraq and how impact it to voters' participation in October 2021.the mean of responses 3.48 from them believed you decision to participate in the electronic vote was correct and this value not encourage. In previous years, if the use of technology in Iraq was done in a specific way, it was limited to some Aspects of the electoral process, and technology was not used in the voting stage, which is one of the most important stages The electoral process because it represents a translation of the will of the people in drawing up the general policy of the country. The importance of this research on the electoral system dependent to electronic voting lies in ensuring its transparency and impartiality and the possibility of its application in Iraq.

The use of Automated Fingerprint Identification Systems (AFIS) for person identity based totally on fingerprints is giant in many countries. However, when dealing with large databases, the sheer variety of fingerprint comparisons required may be overwhelming. To cope with this mission, fingerprint class is hired to reduce the range of necessary comparisons, consequently enhancing the efficiency of database searches. Deep knowledge of strategies, in particular in the field of laptop imagination and prescient, have made large advancements. This paper conducts a comprehensive survey of studies related to fingerprint type the use of deep convolutional mastering. The survey affords an overview of convolutional deep neural networks (CNNs) and highlights various varieties of CNN fashions utilized in fingerprint class tasks. Furthermore, it affords a radical contrast and evaluation of ten study projects in this area, deliberating the CNN models hired, the datasets used, and the carried out accuracy along with the associated hyperparameters. Among the findings, the most general CNN architectures hired in fingerprint category show off high-quality accuracy quotes of 98.9%, 99.2%, or even a 100%. This underscores the effectiveness of deep studying strategies in improving fingerprint-type structures.

These days, it is much of the time tracked down an extensive variety of the getting data. The comprehension that the creator needs to configuration is about cryptography. Cryptography is the investigation of the security of information or data, regularly tracked down in moving information. One of them is the strategy for GOST, GOST is a shortening of "Gosudarstvennyi Standard" or "Government Standard." The calculation is straightforward encryption calculation which has a few cycles upwards of 32 adjusts and utilizes 64-digit block figure with 256-bit key. GOST technique likewise utilizes the S-Box 8 bits of long-lasting and XOR tasks and Pivot Left Shift, the proposed structure creates subkeys utilizing replacement of S-box powerfully and introductory table to gives the presentation of GOST. The move toward thinks about various security perspectives and measurements assessment for confirmation.

Evolutionary or genetic algorithm mimics nature's principles of evolution to find the optimal solution to any problem that is compatible with its mechanism. The genetic algorithm (GA) is one of the most well-known heuristic methods for resolving computing problems. We discover that Darwinian evolution's characteristics have been expressed after studying the GA. It has actually had a number of achievements in various areas of life's demands. The magic square (MS), which is an array of positive integers (1, 2, n) configured so that the sum of the n numbers in any principal diagonal, horizontal, or vertical line is always the same number, was correctly obtained for this study using a GA. In particular, evolutionary algorithms have been utilized as a way to find the best MS solution. How successfully does GA optimization accomplish the MS is the research subject. This study has successfully established a way to build Latin Squares that can adapt to variable sizes (N x N), allowing them to handle Latin Squares of various lengths. This novel method offers a more adaptable method for dealing with Latin Squares of any scale, diverging from the traditional emphasis on specific numbers or defined dimensions.

Object recognition is tough in image processing. The success of many computer vision applications depends on object recognition. For years, object recognition techniques and systems have been developed to solve this challenge. This article reviews computer vision and object recognition methods. In this survey we discuss challenges faced by the computer vision, Objectives of Research in this field ,Types of Architecture in CNN, point detector, frame differencing to detect objects. The article discusses object detecting strategies from several researchers.

The multiple linear regression analysis shows and explains the effects between the explanatory variables and the dependent variable, as its function is to represent the data to understand the nature of the relationship between the explanatory variables and the response variable among the problems that the regression suffers from is that the data does not follow the normal distribution and the emergence of different types of regression models such as the beta regression model, and to know the estimates of the parameters of the multi-linear model, several methods are used for estimation, as this stage is considered one of the important and necessary stages in the analysis of the model, and because some of these methods are traditional and do not give accurate results because the random error does not have a normal distribution, so other methods with good results and suitable for the error distribution are used, so the beta regression model will be studied in which the random errors follow the beta distribution and where the data is smaller than one and it is confined within the period (0,1), the beta regression model will be addressed in this research when there is a problem of multicollinearity, and a comparison will be made between Liu's method and the proposed method used to estimate the model's parameters, as the comparison is done by simulation in two methods.

In the past two years, remarkable technological advancements have led to the unveiling of CHATGPT-3 in November 2022. This cutting-edge creation has garnered immense global demand across diverse demographics. CHATGPT-3 stands as a cornerstone in contemporary artificial intelligence, a synthesis of multiple technologies such as Natural Language Processing (NLP), Machine Learning (ML), and Artificial Intelligence (AI), all harnessed by its parent company, OpenAI functioning as an open-source innovation, CHATGPT-3 draws upon an extensive reservoir of knowledge spanning various fields of science. Through meticulous training on this expansive dataset, the program adeptly furnishes precise responses in manifold domains, encompassing but not confined to education, media, healthcare, sports, and numerous other scientific and humanities disciplines. This progressive evolution culminated in the release of CHATGPT-4, an even more advanced iteration of the technology. The implications of CHATGPT-3 are resounding. It has revolutionized diverse sectors, including media, where it facilitates interactive and informative exchanges with users. In the realm of programming, CHATGPT-3 exhibits its prowess by aiding developers with coding tasks, offering solutions to complex problems. Furthermore, its contribution to scientific research is invaluable, aiding researchers in gathering insights and generating hypotheses. The CHATGPT-3 extends its capabilities to the generation of research papers, a feat that underscores its potential to streamline scholarly pursuits. In the domain of education, this technology emerges as an indispensable tool, providing personalized and comprehensive learning experiences. Its impact thus resonates across academia, professional domains, and beyond.

Injuries resulting from traffic accidents are the first cause of death for persons aged 15 to 29 years, according to the World Health Organization report on traffic accidents issued on 20-June-2022. For the study and analysis of traffic accidents, it is preferable to use spatial analysis to know the influencing causes and according to each region. Therefore, geographically weighted Poisson regression models (which is an updated model of the traditional regression, so that the diagnosis is more accurate for the studied phenomenon by adding the spatial dimension to the model, so the new model takes into account this relationship (the spatial dimension), that is, the geographically weighted regression ( GWR) performs a per-site regression rather than a single regression for the entire study area) were used for this purpose. As the fixed relationship may hide some important spatial factors that affect the relationship between the response variables and the explanatory variables, and then may negatively affect the accuracy of the models. The geographically weighted Poisson regression (GWPR) model was used in this study on a sample of traffic accidents for the period from 2006 to 2022 for all governorates of Iraq except for the Kurdistan region for the purpose of carrying out prior work to determine the locations in which traffic accidents may increase. Two Kernel functions (gaussian and tri-cube) were also used. The study showed that the tri-cube function was the best and according to the corrected Akaike information criterion (AICc) .

Grey system theory is regarded as a multidisciplinary scientific approach, which deals with systems that have partial information (such as: small samples). Grey modeling as a vital portion such theory that yields a sufficient results despite of limited amount of data. Grey Models can be divided into two types; univariate and multivariate grey models. Univariate grey model GM (1,1) is the cornerstone of this theory, but it doesn’t take the relative factors in account. While multivariate grey models GM(1,M) takes those factor in account. However, they have a complex structure and some defects in " modeling mechanism", "parameter estimation "and "model structure". Therefore, GM(1,M) in the traditional version was submitted to many trials of optimizations to getting rid this defects.
This paper aims to:
•
introduce the characteristics of the traditional GM(1,M) as well as its problem.
•
impose a method to overcome such problem, through two optimized multivariable grey model of one order derivative equation.
•
present an Optimized Grey Model abbreviated as OGM(1, M) by adding the linear correction term h1(M-1)and greyaction quantity term (h2) to the traditional model GM(1,M).
•
use an Optimized Background value Grey Model OBGM(1,M) by optimizing the Background value of the last model OGM(1,M).
•
apply real data-set represents water consumption in Baghdad at the period (2016-2022) to compare both optimized models with the traditional model.
•
use the mean absolute percentage error (MAPE) and the determination coefficient R2.
The results of comparison showed that both OGM(1, M) and OBGM(1,M) have superiority compared with the traditional model GM(I,M).

Data gathered from the observation of a phenomenon are not all equally informative some of them may be noisy, correlated, or irrelevant. Variables selection aims at selecting a variable set that is relevant for a given task. This problem is complex and remains an important issue in many domains. This paper is a review of Genetics Algorithm (GA) as a variable selection method for the multiple linear regression model. Stepwise regression had been used as a traditional method for the selection of the best subset of variables to be included in the model. Adjusted R squared along with Akaike Information criterion (AIC) and Bayesian Information criterion (BIC) are used as a model’s fit criteria. The research encompassed daily monitoring of nine variables, which included global horizontal irradiance, air temperature, relative humidity, dew point, wind speed, wind direction, precipitable water, azimuth, and zenith. The study sample was stratified by time into two seasons, namely the hot and cold seasons. The dataset for the hot season comprised 2571 readings, while the cold season dataset consisted of 2075 readings, encompassing the timeframe from January 1, 2021, to December 31, 2021, for Baghdad. The findings demonstrated the efficacy of the genetic algorithm approach in selecting model variables based on evaluation criteria (AIC, BIC, and adjusted R squared) surpassing the performance of stepwise regression.