ABSTRACT: 

Facial categorization and visual recognition of unknown gunmen in Nigeria using machine learning algorithms has emerged as one of the greatest tools in the field of artificial intelligence and technology to detect and identify unknown gunmen with painted and masked faces. Apparently, it brings together the conglomeration of various technologies such as feature extraction, data normalization and the data training of the images in an attempt to identity individuals (unknown gunmen) with distinctive features of the various parts of the body. This is done through a more complex surveillance camera system or commonly used handsets and webcam technologies to capture the images of the individuals in question. The machine learning algorithm employed in this work is YOLO (You Only Look Once). The YOLO was employed in this work to create a binary classification, classifying the data into unknown gunmen and not gunmen and the model of the facial recognition and visualization of unknown gunmen were created. The application for the system was developed using python programming language. Confusion matrix was used to test the performance of the algorithm using the instance from 0 to 80. The accuracy of the model was determined through the precision recall and it yielded 0.995 for the not gunmen and 0.958 for the unknown gunmen.

Keywords: unknown gunmen, machine learning, categorization, visualization.

ABSTRACT: 

This study explores the causal relationship between crime investigation methods and their impact on investigation outcomes. Key factors under scrutiny include the duration of investigations, case discharges and acquittals, convictions, and investigation efficiency. Data was gathered via a survey encompassing 184 participants from six security and law enforcement agencies in Nigeria. Analysis was conducted employing descriptive statistics and the statistical independent-samples t-test within SPSS. Findings indicate a limited integration of digital forensics (DF) within crime investigation practices in Nigeria. The t-test underscores significant disparities between the DF-utilizing group and the non-utilizing group. Notably, substantial variations were observed in investigation duration (t181.614 = -7.098, p < .005), instances of case discharges due to investigative errors (t176.106= -8.385, p < .005), conviction rates (t181.960= 14.848, p < .005), and investigation efficiency (t176.331= -7.138, p < .005). Consequently, embracing appropriate digital forensic techniques is pivotal for enhancing the efficacy of crime investigation in Nigeria. 

Keywords:  unknown gunmen, machine learning, categorization, visualization.

ABSTRACT: 

This study investigates the performance of senior secondary school students in mathematics at Uyo High School over the past decade (2013-2022) using Support Vector Machine (SVM) analysis. The research aims to identify patterns and trends in student performance, exploring the impact of demographic factors such as gender, age, ethnicity, and socioeconomic status on mathematics achievement. Using a quantitative research design, the study gathers and preprocesses mathematics senior secondary exam results from Uyo High School during the previous ten years. The dataset was subjected to extensive preparation, which included handling missing values, encoding categorical variables, and scaling numerical characteristics. It included demographic data and mathematical performance ratings. Feature selection approaches were used to identify the most important demographic indicators for predicting math achievement. The SVM classifier was trained using demographic variables to predict mathematics performance categories. Accuracy, precision, recall, and F1-score metrics were used to assess the model. The findings provided important direction for specialized interventions and instructional tactics to improve student outcomes by illuminating the association between demographic characteristics and mathematical ability. To improve students' performance in mathematics, the study ends with recommendations for instructional tactics and interventions. By offering insights into the academic achievement of senior secondary school pupils at Uyo High School during the previous ten years, the study adds to the corpus of research already available on mathematics education.

Keywords:  Support Vector Machine, Mathematics Performance, Senior Secondary School, Examination Results, Instructional Strategies, Interventions.    

ABSTRACT: 

Wind energy, a renewable resource, supports sustainable energy production without environmental pollution. Numerous studies aim to maximize wind energy capture. One approach involves boosting wind turbine efficiency by increasing wind speeds. This is achieved through a turbine-concentrator system, where a concentrator surrounds the turbine, creating a low-pressure zone behind it to accelerate wind speed at the turbine. Experimental studies analyze various combinations of turbines and concentrators to optimize wind speed enhancement. This study introduces a novel concept: Straight-bladed Vertical Axis Wind Turbines (SB-VAWTs) equipped with convex-shaped wind concentrators. These concentrators are positioned above and below the rotor to enhance airflow capture and improve internal flow dynamics. Concentrator dimensions are tailored to SB-VAWT specifications to study their combined aerodynamic performance through experiments. Results indicate that integrating wind concentrators significantly enhances SB-VAWT aerodynamics, especially at lower wind speeds. In this study, the average power coefficient of SB-VAWTs equipped with concentrators increases by up to 41%. These findings underscore the potential of wind concentrators in enhancing SB-VAWT efficiency and suggest their application in wind energy systems.

Keywords:  Traight-Baded Vertical Axis Wind Turbine (SB-VAWT); Wind energy; Wind concentrator; Aerodynamic performance.

ABSTRACT: 

The paper proposes an intelligent-based design to Data Security Monitoring in a Cloud Computing Infrastructure through the application of enhanced and Robust Virtualization Model. This technology becomes eminent due to the simultaneously alarming Malicious activities which is the major issue in cloud computing. When we talk about cloud computing, we refer to a model that enables convenient, on-demand network access to a shared pool of configurable computing resources such as networks, servers, storage, applications and services, etc. This can be rapidly provisioned and released with minimal management effort or service- provider interaction. In this paper, a virtualized model was developed to enhance data security monitoring. The Structured System Analysis and Design Methodology was adopted and design was achieved through the use of such tools as Dataflow Diagram, Use-Case Diagram, Unified Modelling Language (UML) Diagram, and Sequence Diagram. Five Hundred (500) datasets was used from robust repositories, of which 30% was used for training, while 70% was used for testing. Parameters for analysing and evaluating the results of both systems encompassed the number of adopted algorithm, the number of adopted technologies, the number of adopted design tools and the number of tested records. From the performance evaluation, the new system showed better performance than the existing system as it achieved an accuracy rate of 1.07% as compared to the existing system which achieved an accuracy rate of 0.48%. The newly developed model was for fraudulent data detection in cloud computing infrastructure with special emphasis to financial fraud. This is because; financial frauds are committed against property, involving the unlawful conversion of the ownership of property to one's own personal use and benefit. In addition, the new system was further optimized with deep neural network and logistic regression technique. This study could be beneficial to anti-corruption agencies, corporate organizations and researchers with keen interest in the study area.

Keywords: Memory virtualization, Nested virtualization, Distributed file system, logistic regression technique, and Cloud Computing.

ABSTRACT: 

This study presented an optimized approach to Data Security Monitoring in Cloud Computing Infrastructure via an Improved Robust Virtualization Model. Malicious activities have continued to become an alarming issue in cloud computing. Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service-provider interaction. In this work, a new virtualization model for data security monitoring was developed. Structured System Analysis and Design Methodology was adopted and design was achieved with tools such as Dataflow Diagram, Use-Case Diagram, Unified Modelling Language (UML) Diagram, and Sequence Diagram. The study utilized Five Hundred (500) datasets from robust repositories, of which 30% was used for training, while 70% was used for testing. Parameters for analyzing and evaluating the results of both systems encompassed the number of adopted algorithms, the number of adopted technologies, the number of adopted design tools, and the number of tested records. From the performance evaluation, the new system showed better performance than the existing system as it achieved an accuracy rate of 1.07% as compared to the existing system which achieved an accuracy rate of 0.48%. The newly developed model was for fraudulent data detection in cloud computing infrastructure with a special emphasis on financial fraud. This is because; financial frauds are committed against property, involving the unlawful conversion of the ownership of the property to one's personal use and benefit. In addition, the new system was further optimized with a deep neural network and logistic regression technique. This study could be beneficial to anti-corruption agencies, corporate organizations, and researchers with keen interest in the study area.

Keywords: Logistic Regression Technique, Virtualization, Cloud Computing, Security Monitoring, Virtualization Architecture.

ABSTRACT: 

Medical images are a special kind of images. These images are used to help doctors diagnose diseases in patients. There are many methods for capturing these images using modern technology and specialized cameras, microscopes, and other tools. This study focuses on images in the BMP 24 RGB color format (8 bit for RED, 8 bits for Green, and 8 bits for Blue). Many disease states, especially those involving blood diseases, have difficult diagnostic processes, or the patient can need several tests or a long period of time to be diagnosed. In cases of leukemia and for the purpose of diagnosing blood cells, for example, such operations make it easier to determine the shape, size, area of distribution of red or white blood cells, and other physical characteristics. It is simple to decide these physical features as the binary image or contains two colors, one of which represents the cell and the other the plasma in the blood. In this study, we will convert medical images into binary images in order to make it easier for the doctor or other smart electronic software used for diagnosis to work with them.

Keyword:  Binary Images, diagnosis, RGB, Grayscale, Image Processing.