Pusat Pengajian Sains Matematik - Tesis
Browse
Recent Submissions
- PublicationNew Generalized Differential Transform Method For Solving Fractional Ordinary Differential Equations(2025-06)Abuualshaikh, Ammar Jamil AhmadFractional differential equations (fdes) have garnered significant attention in recent years due to their ability to model nonlocal and memory-dependent phenomena in various scientific and engineering domains. Unlike classical differential equations, fdes involve derivatives of non-integer order, leading to intricate and novel mathematical properties. This unique characteristic has allowed fdes to find successful applications in fields such as control theory, chemistry, economics, physics, finance, medicine, and biology. Consequently, there is a pressing need for innovation and the development of accurate and efficient solution techniques for these types of differential equations. However, one of the main challenges in dealing with fdes is the scarcity of exact solutions, mainly due to the complex nature of the real-world problems they represent. To address these challenges, the main goal of this thesis is to introduce a novel method called the new generalized differential transform method (ngdt). The ngdt method utilizes the riemann-liouville derivative (rld) and is designed to offer analytical or numerically approximated solutions for particular instances of ordinary fdes, whether in linear or nonlinear form.
- PublicationThe Impacts Of Transiting To Green Energy In Africa: Evidence From Gis Mapping And Spatial Panel Models(2025-05)Gwani, Alhaji AbdullahiThis thesis examines the spatial dynamics of renewable energy production and consumption (rep and rec) and their impact on economic growth (gdp), ecological footprint (ef), and carbon dioxide emissions (co2) in 49 african countries over the course of three decades (1990-2020). The study underlines the necessity of a transition to sustainable green energy in light of the urgent environmental challenges associated with conventional energy sources (fossil fuel). The objectives are as follows: (1) to analyse the trends in rep and rec, (2) to explore the spatial autocorrelation and dynamic characteristics of data, (3) to investigate the impact of rep and rec on economic growth and environmental degradation, and to identify the most effective spatial panel model for these dynamics. To overcome the estimation issue in the conventional non-spatial model that fails to consider spatial correlation in the data, the gis mapping and spatial panel models are utilised in this study. The results are compared with the non-spatial model. Gis mapping shows that ef and co2 emissions experienced a decline, particularly between 2017 and 2020, indicating a positive trend toward greener energy and environment while both rep and rec demonstrate consistent growth.
- PublicationMutation Tabu Search Using Systematic Probabilistic Two Satisfiability In A Discrete Hopfield Neural Network(2025-09)Chen, JuSatisfiability logic has been extensively applied in artificial neural network research. However, existing studies have not developed effective strategies for controlling the distribution of literals in the formulations. Therefore, this thesis proposes a Probabilistic 2 Satisfiability model that controls the quantity and position of positive literals through a positive rule applied to variables in second-order clauses.
- PublicationOptimization Seaweed Drying Efficiency Using Hybrid Solar Dryers And Sparse Robust Regression Models(2025-04)Afouna, Nour Hamad Suleiman AbuData analytics in statistics is vital for extracting insights, identifying patterns, and guiding decisions. In precision farming, particularly post-harvest management, challenges arise from iot sensor dependency, system complexity, and variable interactions, leading to issues like variability, multicollinearity, and sensitivity to outliers. Addressing these challenges requires improved data inclusivity, robust data management, and cross-sector collaboration to unlock the full potential of analytics. Variability in agricultural systems impacts crop yield and post-harvest processes. Heterogeneity in sensors, data collection methods, and transmission protocols complicates agricultural drying. Multicollinearity, where independent variables are highly correlated, creates difficulties in post-harvest monitoring as overlapping environmental data from multiple sensors obscures the impact of individual variables. Fluctuations due to environmental changes, sensor errors, and human interventions further complicate modeling, requiring robust statistical methods capable of handling noise and outliers.
- PublicationAnalysis Of Nonlinear Dynamics Of Epidemiological Models With Local Dispersal, Reinfection And Limited Medical Resources(2025-09)Salman, Al Zaidi Amer MohammedExamining the dynamics of disease spread and its control measures remains a main challenge in epidemiology. Numerous studies have proposed various factors influencing disease transmission, including environmental conditions, population movement, and healthcare resources. While distinct models have been formulated to examine these factors, less attention has been given to understanding how the interplay of reinfection, medical resource limitations, and spatial dispersal processes determines epidemic outcomes. This thesis investigates the long-term dynamics of COVID-19 in Malaysia using a group of deterministic Susceptible-Infectious-Removed (SIR) kinetic models that incorporate temporary immunity, spatial heterogeneity, local dispersal, and control measures.