Doctoral

Permanent URI for this collection

https://repository.futo.edu.ng/handle/20.500.14562/1260

Browse

Browsing Doctoral by Title

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Hybrid deep learning-based model for covid-19 prediction and interpretation using multiple data modalities
    (Federal University of Technology, Owerri, 2024-05) Dokun, Oyewole
    This research addresses the critical need for accurate and timely COVID-19 diagnosis and prognosis by developing a hybrid deep learning model that integrates multiple data modalities, including chest X-rays, Computed Tomography (CT) scans, blood smears, and clinical data. The model employs specialized architectures such as Residual Network with 50 Layers (ResNet50) for Chest X-ray, InceptionV3 for CT scans, Convolutional Neural Network (CNN) for blood smears, and a Random Forest classifier for clinical data analysis. The results demonstrate high accuracy rates: 96.7% for ResNet50, 97.58% for InceptionV3, 96.12% for CNN, and 98.30% for the Random Forest classifier. Grad-CAM enhances transparency by visualizing critical regions in the images, aiding healthcare professionals in understanding the model's decisions. This hybrid model offers improved accuracy and reliability for COVID-19 diagnosis and prognosis, making it a valuable tool for clinical settings and resource allocation. The research underscores the potential of multi-modal data integration in medical AI and suggests further exploration and refinement of such models for broader healthcare applications.
  • Thumbnail Image
    ItemOpen Access
    Intelligent evaluation system for software quality measurement
    (Federal University of Technology, Owerri., 2022-12) Nwandu, Ikenna Caeser
    The concern about the large-scale and complexity of contemporary software cannot be over-emphasized. This is inclined to the assurance of standardized software quality which is essential for preventing disastrous effects of releasing fault-prone systems. This thesis designed an intelligent model that uses various metrics corresponding to six quality attributes (namely Reliability, Usability, Efficiency, Functionality, Maintainability and Portability) to measure the quality of software. This agreed to the assertion that software quality evaluation process is an instrument that observes the characteristics of a software product. In software engineering, the primary quality evaluation and assurance technique that establishes confidence over successful execution of software is termed software testing. Software testing usually identifies and applies metrics to software products in order to promote and assess their quality. This thesis designed an intelligent evaluation model in conformance with software testing principles. The objective of the model is to apply reinforcement learning in its software evaluation process to measure six software attributes in terms of speed of execution and to ensure optimal decision-making in the evaluation process, such that the model returns a reliable outcome. The model utilized a formulated model equation, whose input are the measured attributes, to achieve the evaluation. The model is developed using extreme programming principles, an agile framework whose operation is based on simplicity. It also adopted object-oriented analysis and design methodology which allowed the utilization of various artifacts including use cases, data flow, sequence, flowchart, entity-relationship and class diagrams to describe the architecture and functionality of the system. The model was implemented using Python programming language with the database design on MySQL platform. The model is further validated by comparing its performance measures on test data gotten from the functional information of Oil-palm Management Program and Estate CanePro. These tests data produced quality values of 0.9 and 1.0 respectively via the model equation. These results gave the indication that the resource software perform efficiently owing to the fact that the model’s value benchmark is best as it approaches unity. The result of comparing the outcomes showed that reinforcement learning makes software evaluation dynamic and precise. The results indicated that the model independently determines the strategies to follow during evaluations and the same set of data consistently gives the same outcome. The result also showed that the reliability of a software is directly proportional to its usability and maintainability. However, the result also showed that having a high portability value does not guarantee the reliability and/or maintainability of a given software.
  • Thumbnail Image
    ItemOpen Access
    Modelling of Nigeria’s Liquefied Natural Gas Shipping Trade
    (Federal University of Technology, Owerri., 2022-12) Igboanisi, Chinaemerem C.
    Nigeria has the largest proven natural gas reserves in Africa and its reserves ranked as ninth (9th) largest in the World- accounting for 188.8tcf (trillion cubic feet) of proven reserves as at the year 2019. However, Nigeria’s capacity to participate in the global natural gas shipping trade and earn freight revenue has been constrained by shipping tonnage market domination by other nations. Thus, as the nation strives to improve her revenue earnings through robust visible and invisible trade policy; it has become imperative to investigate empirically the determinants of Nigeria’s international shipping trade in Natural gas. This research developed the gravity model of Nigeria’s natural gas (NLNG) shipping trade to determine the factors affecting NLNG international freight market. The secondary data for the study comprised of volume of natural gas production (in billion cubic meters) shipped between Nigeria and other trading partner countries, geographical distance data between trading partner countries, population mass of trading partners, price of natural gas and bilateral trade agreements. Others include: logistics performance indices and shipping freight rates. These were sourced from global databases, Nigeria LNG limited, the Nigerian Ports Authority and covered the periods between years 2003 to 2020. To address the hypotheses governing this research, we developed an augmented gravity model of natural gas shipping trade in Nigeria’s international freight market and examined trends in demand. The following variables were found statistically significant in explaining NLNG trade namely: quality of transport infrastructure (-225.448), geographical distance (-232.721), trade agreement (42.534) and population mass (0.955). These coefficients are in their natural logs and can therefore be interpreted as elasticities. In terms of most important trading blocs or shipping routes, the most important shipping routes (which are dummy variables) are namely: The United States of America (3,360.056), EuroAsia (3,090.082), Europe (904.810) and South America (786.413). These findings indicate that robust policy interventions are needed to promote trade with our trading partners. Robust investments are also needed in our transport infrastructure quality (especially that of bunkering facilities for LNG vessels) in order to reduce impediments to trade. From the positive trend analysis results, demand for natural gas is positive and the federal government should encourage more private sector investment in LNG shipping fleet to increase Nigeria’s participation in LNG international freight market. As recommendation for further studies, modelling of constraints of natural gas trade involving gasification and re-gasification stations should be explored in order to expand the scope of the present work.