CENTRAL ASIAN JOURNAL OF MATHEMATICAL THEORY AND COMPUTER SCIENCES

E-School Information Management System (E-SIMS)

2025-04-06T16:17:07+00:00

In order to appropriately and effectively address all of the long-standing school information problems in Iraqi schools, this study aims to build and use a requirement model as a foundation to develop an electronic school information management system (e-SIMS). Online learning and related activities were the subject of this case study, which looked at elementary schools in Iraq. It is my sincere wish that the system developers will find this model useful in gaining a better grasp of the needs for the e-school activity management system's concepts, processes, and procedures. In this study, the requirements were defined using fact-finding techniques such as observation, interviews, and requirement model analysis. This requirement model, which includes specific model diagrams, was built using Unified Modeling Language (UML). The requirement model for e-SIMS was built using a combination of visual aids like use case diagrams, class diagrams, activity diagrams, and interaction diagrams (sequence diagrams and collaboration diagrams). The model was backed up by textual information such as a use case specification and a requirements list with 51 functional and non-functional requirements. Nevertheless, the scope of this investigation is limited to capturing functional requirements. The test script technique and the prototype system were used to validate this concept. e-SIMS is an online platform that aims to improve communication between users, regardless of their location or the time of day. Findings from this research suggest an improved approach to creating e-SIMS that relevant education organizations in Iraq can use.

Predicting The Number of Traffic Accidents Using Box-Jenkins Models to Contribute to Reducing These Accidents (An Applied Study in Baghdad Governorate)

2025-04-07T00:54:50+00:00

This research aims to analyze the time series of the number of recorded accidents in Baghdad during the period using Box-Jenkins models to find the best and most efficient predictive model for the number of accidents during the period. The results، based on comparison criteria (AIC، SCH، HQC) for the significant models and the comparison between the proposed parameters and models، showed that the suitable model for estimating the number of accidents is the ARIMA(1،1،2) model. The predictive values have shown consistency with their counterparts in the time series with the actual values in the trend، indicating the efficiency of the model.

Error Analysis and Stability of Numerical Methods for Solving Fractional Differential Equations in Biophysical Modeling

2025-04-09T02:00:19+00:00

Fractional differential equations (FDEs) have eCombined as a powerful tool for Representationing Complicated biophysical phenomena such as anomalous diffusion and viscoelastic behavior due to their ability to capture memory effects and hereditary properties. notwithstanding reAnswer fdes numerically presents important challenges including Problems of truth constancy and computational Productivity. This paper addresses these challenges by proposing and analyzing a novel numerical method tailored for solving FDEs in biophysical contexts. the wise employs amp limited limited Disagreement access with accommodative time-stepping ensuring both great truth and constancy spell maintaining computational feasibleness. A rigorous theoretical analysis is conducted to establish error estimates and stability conditions demonstrating the method consistency and convergence properties. quantitative experiments are performed along pragmatic biophysical problems such as arsenic abnormal dissemination inch tProblems and elastic matter distortion to corroborate the method operation. The results show that the proposed scheme achieves first-order temporal Precision and second-order spatial Precision outperforming standard techniques like the Grünwald-Letnikov method in terms of both precision and Productivity. Furthermore, the wise exhibits iron constancy low variable down orders and measure sizes devising it good for long Imitations of biophysical systems. These findings underscore the potential of the proposed approach to advance our understanding of Complicated biological Methods and Improve Foretelling Representation Ing in biophysics. away addressing name limitations of present methods this read Adds to the evolution of true and prompt quantitative tools for reAnswer fdes inch pragmatic Uses.

Modeling the Evolution of Population Dynamics Using Ordinary Differential Equations: Mathematical Analysis and Modern Applications

2025-04-11T02:29:51+00:00

Population dynamics modeling plays a critical role in understanding ecological stability, epidemiological spread, and sustainable resource management. Classical models such as the logistic growth and Lotka-Volterra equations offer foundational insights but often overlook environmental stochasticity and multispecies complexity. Existing frameworks frequently simplify nonlinear feedbacks or exclude real-time ecological interactions, limiting predictive capacity in dynamic systems. This study advances population modeling by integrating ordinary differential equations (ODEs) with modern computational tools, including machine learning-enhanced simulations and neural ODE frameworks. Analytical techniques such as linear stability, Lyapunov methods, and bifurcation analysis revealed equilibrium classifications and transitions in logistic, predator-prey, and epidemiological models. Numerical simulations validated theoretical findings, showing that hybrid AI-augmented models achieved higher accuracy (relative error 2.1%) and computational efficiency (98%) than traditional models. A novel contribution lies in embedding neural networks within classical ODE systems to dynamically adjust model parameters using heterogeneous data streams. The developed framework enhances the realism and adaptability of population models, with direct applications in conservation planning, disease control, and ecological forecasting, thus offering a versatile and interdisciplinary tool for addressing real-world biological challenges.

A Review on Development of a Client-Server Chat Application Using Python for Real-Time Communication and Networking

2025-04-13T04:55:08+00:00

This paper focuses on developing a client-server chat application using Python to explore key concepts in computer networking and real-time communication. The client-server architecture acts as the foundation, with the server managing message routing and connections between multiple clients. Using TCP/IP protocols, the system ensures reliable data transfer, error checking, and segmentation. Python's socket programming facilitates connections, where sockets serve as endpoints for data exchange. The server listens for incoming connections and employs multi-threading to handle multiple users concurrently without affecting performance. The chat application emphasizes network security by considering encryption mechanisms like SSL/TLS to safeguard communication. Additional security measures such as error handling, timeout mechanisms, and user authentication can further enhance its robustness. The paper demonstrates essential networking concepts like packet transmission, client-server interactions, and data flow management. Its real-time nature ensures seamless information exchange, improving user experience. Future enhancements include message broadcasting, private messaging, and a graphical user interface (GUI) for better usability. The paper serves as a foundation for understanding modern messaging systems and can evolve with advanced security protocols and additional functionalities like file sharing. Overall, it provides a practical learning experience in network-based application development and security implementation.

Developing an Inclusive and Efficient Extractive Text Summarization System Using NLP Algorithms

2025-04-15T04:29:18+00:00

From brainstorming to code coding and beyond, the research covers it all in its comprehensive analysis of text summarisation system development. Improving accessibility and user experience in the field of extractive text summarisation is its key objective. The study's overarching goal is to create a novel summarisation model that can employ natural language processing (NLP) to efficiently condense information while also meeting the varied requirements of its users. This involves using cluster, graph-based ranking algorithms for Natural Language processing, and TextRank to facilitate the extraction of important information and the development of short summaries. To ensure that people with different abilities can make full use of the summarisation system, the research also takes a user-centric approach, which emphasises accessibility. Therefore, we work to meet a variety of accessibility needs, such as providing alternative formats for visually impaired users, including voice interfaces for motor impairment users, and introducing future features like adjustable reading speeds and screen reader compatibility to allow for user preference-based customisation. We have utilised ROUGE-L for testing and evaluation. An inclusive and powerful text summarisation system that caters to users' demands and improves their experience is the ultimate goal of this research.

Development of a Hybrid Optimization Algorithm for Efficient Neural Network Training

2025-04-26T03:38:27+00:00

In this study, a hybrid optimization algorithm in mathematical combining the Hestenes-Stiefel (HS) and Polak-Ribiere (PR) methods was developed using an adaptive approach to optimize the training of Feed-Forward Neural Networks (FFNNs). This approach aims to leverage the global convergence power of the HS method and the ability of PR to avoid local minimum. The hybrid algorithm was tested on three real world datasets (Iris, Glass, and Wine), and compared to the results obtained using the original two methods (HS and PR). The hybrid algorithm showed shorter training times across all datasets and hybrid algorithm significantly reduced the mean square error (MSE) compared to the two separate methods, resulting in faster convergence with training accuracies fir the Iris datasets, Glass dataset, and Wine dataset being 98.50%, 98.39%, and 65.98%, respectively, enhancing efficiency. The algorithm also proved to be able to handle data with high variance or nonlinearity more effectively, making it suitable for training neural networks in machine learning applications.

Real-Time DDoS Attack Detection using Random Forest and Streamlit: A Machine Learning-Based Web Solution

2025-04-20T17:41:35+00:00

DDoS assaults threaten network availability and security in today's linked world. We propose a real-time DDoS assault detection method using machine learning, specifically the Random Forest algorithm. Our technology fits smoothly into web settings using the renowned Streamlit architecture, giving an easy and interactive platform for threat monitoring and mitigation. We collect a large dataset of network traffic features from benign and harmful operations. After careful preprocessing and feature engineering, we prepare data for training and evaluation. We use the resilient and scalable Random Forest method to create a predictive model that can distinguish typical traffic patterns from DDoS attacks. The model is rigorously tested using performance indicators to detect and categorize DDoS assaults with low false positives. The model is effortlessly integrated into a Streamlit-based online application, improving end-user accessibility and usability. Our experiments show that the model can identify in real time with excellent accuracy and efficiency. Our technology strengthens network resilience and protects dynamic online environments from disruptive cyber threats by empowering stakeholders with proactive DDoS mitigation capabilities.

Utilizing Fuzzy Linear Regression in Medical Data Analysis: A Comparative Study of FLR and FLSLR Models for Enhancing Regional Health Planning

2025-04-23T21:26:29+00:00

This research deals with the planning study in using fuzzy linear regression in the cases of fuzzy; non-fuzzy data, where the analysis was conducted using several methods, including fuzzy linear regression (FLR) and modified fuzzy linear regression, in counting up to the (FLR) method using least squares (FLSLR), in which linear programming (LP) was employed in the analysis process.The fuzzy parameters were estimated based on fuzzy and non-fuzzy data, and the regression idol was determined based on the concepts of fuzzy set theory. These methods were also applied in the medical field to accurate data on osteoporosis in a statistical study as part of the requirements for achieving regional planning, which is concerned with health population studies as an integral part of sustainable development, in which individual health is one of the most important variables affecting planning levels in general, where the researchers used this method, by measuring bone density using a DEXA device for thirty patients (10 males and 20 females). The study was conducted in the laboratories of the Department of Biomedical Engineering - College of Engineering - Al-Nahrain University.The results showed that the modified Tanaka model is more efficient than the FLR model, as it helps to elude the emergence of non-fuzzy given parameters. The FLSLR method also proved superior to FLR based on the degree of model affiliation, which enhances the accuracy and effectiveness of the estimation.

The Weighted Time-Domain Laplace-Fourier Transform: A Comprehensive Study

2025-04-24T17:51:07+00:00

Traditional integral transforms like Fourier and Laplace have been foundational tools for signal and system analysis in physics, engineering, and mathematics. However, these classical methods are often inadequate for analyzing nonlinear, non-stationary, and time-varying signals that are increasingly prevalent in real-world systems. In response to these limitations, this study introduces the Weighted Time-Domain Laplace-Fourier Transform (WTLFT), which integrates the benefits of Fourier and Laplace transforms with a time-domain weighting function to enhance analytical flexibility. Prior methods lack robustness in capturing transient behaviors and localized features in dynamic systems. There is limited development of transform techniques that address both fractional dynamics and time-varying structures within a single framework. This research aims to define the WTLFT, prove its core properties, and demonstrate its effectiveness in solving complex differential equations involving non-integer derivatives and dynamic systems. The WTLFT is analytically validated through properties such as linearity, time-shifting, convolution, and differentiation. It is successfully applied to exponential and Mittag-Leffler functions, and solves Caputo-type fractional models, diffusion equations, and differential-algebraic systems. Unlike conventional transforms, the WTLFT allows for flexible weighting strategies that adapt to specific application domains, enabling enhanced signal analysis and accurate reconstruction of solutions. The WTLFT establishes a powerful new direction for signal processing, control theory, and fractional calculus applications, especially in engineering, biomedical, and physical science fields.

Gathering of Unit Disk Robots

2025-04-25T15:22:01+00:00

The gathering problem in distributed robotics involves coordinating autonomous mobile robots to converge at a single location, a task complicated by physical constraints and limited sensing. Prior research predominantly utilizes point robots or assumes full visibility, which neglects real-world conditions where robots have physical dimensions and obstruct each other's vision. No existing algorithm in the classical model addresses gathering of unit-disk robots under obstructed visibility using small, constant memory. This study proposes a novel algorithm for collision-free gathering of opaque unit-disk robots with limited memory, operating under a fully synchronous scheduler. The algorithm proceeds in two phases: a visibility phase to reposition robots into a convex hull ensuring mutual visibility, and a gathering phase to move all robots toward the centroid. It achieves collision-free convergence in O(n) rounds using only O(1) memory per robot, without knowledge of the total number of robots. Unlike prior models that rely on lights, transparency, or global knowledge, this work uniquely solves the gathering problem using opaque robots with geometric reasoning and no communication. This result advances the feasibility of memory-efficient, scalable coordination strategies for real-world swarm robotics where physical and visibility limitations are critical.

Expectile Regression with Reciprocal Lassopenalty

2025-04-26T04:19:23+00:00

Expectile regression (ER) has emerged as a valuable alternative to quantile regression, offering robust modeling of the conditional distribution of response variables across diverse fields such as economics, medicine, and ecology. While traditional penalized ER models like SCAD, elastic-net, and adaptive Lasso improve variable selection, they often struggle with achieving optimal sparsity and precision. Existing penalties typically exhibit continuity and symmetry, which may not sufficiently penalize near-zero coefficients in high-dimensional data, limiting their effectiveness in variable selection. This study proposes a novel penalized expectile regression method using the reciprocal Lasso (rLasso) penalty, which introduces discontinuity at the origin and infinite shrinkage as coefficients approach zero. Simulation studies across various settings and error distributions demonstrate that ER-rLasso consistently achieves the lowest root mean square error (RMSE) and false positive rate (FPR), outperforming established ER models. In real-data analysis involving prostate cancer, ER-rLasso showed superior predictive accuracy across all tested expectile levels. The distinct penalization structure of rLasso, being non-symmetric and non-convex, introduces a more aggressive variable selection mechanism that enhances model sparsity and interpretability.

Poisson Distribution and its Relationship to The Normal and Binomial Distributions: Review Article

2025-04-26T04:27:08+00:00

The Poisson Distribution (PD) is a foundational statistical model widely utilized in probability theory to represent the frequency of discrete, independent events within a fixed interval of time or space. Its analytical structure, based on the parameter λ, allows it to effectively model rare occurrences in diverse fields such as telecommunications, health, commerce, and environmental science. It also serves as a mathematical bridge between the Binomial distribution under Bernoulli trials and the Normal distribution under large-sample conditions. Despite the PD’s established applications, a comprehensive synthesis of its convergence behavior and comparative properties with the Binomial and Normal distributions remains underexplored in the literature. This article aims to review the mathematical relationships and convergence properties of the PD, particularly in the context of approximations to Binomial and Normal distributions, and to reaffirm its applicability in modeling real-world phenomena. The analysis confirms that the PD approximates the Binomial distribution when the number of trials is large and the success probability is small, and it converges to the Normal distribution as λ increases. These findings are substantiated by theoretical derivations and supported with examples from current scientific applications. The study unifies theoretical derivations and practical illustrations, highlighting the central role of PD in statistical modeling and its versatility across various domains. The findings reinforce the importance of PD as a core tool in applied statistics and suggest potential for its enhanced application in complex systems through hybrid models and extensions involving fuzzy logic.

Advanced Image Processing for Breast Cancer Detection Using CNN-Based Transfer Learning on Mammograms

2025-04-28T12:47:17+00:00

Breast cancer remains the most commonly diagnosed disease and the second leading cause of death among females. Statistically speaking, roughly one out of every eight American women was diagnosed with breast cancer last year. The precise identification of breast cancer also largely relies upon careful analysis of medical images. Though several Deep Learning (DL) algorithms have been employed to analyses such images, therefore, this study focuses on using a Convolutional Neural Network (CNN) to differentiate between different types of mammograms. The use of CNN in image recognition and visual processing has quickly drawn the attention of scholars. Therefore, in this current research, an approach is presented to extract patches from mammograms and utilize them to train the CNN, whereby the order of the section’s feeds into the classification process. In addition, a transfer learning approach is utilized, in which a model created in the initial phase is later utilized as an initial model. Besides using single and multi-CNN and Artificial Neural Network (ANN) layers, two more approaches—Auto-Encoder and VGG16—are used to evaluate and compare the effectiveness of the models on different datasets.

Exploring Numerical Methods: Solving Lane-Emden Type Equations with Padé Approximations

2025-05-07T16:40:27+00:00

This research explores the effectiveness of using Padé approximations to enhance the accuracy of numerical solutions for Lane-Emden type differential equations. By applying the Adomian decomposition method to series solutions derived from previous studies, Padé techniques are integrated to obtain more precise approximate solutions. The supplied examples demonstrate that Padé approximations extensively outperform conventional strategies, yielding numerical results with smaller mistakes and nearer proximity to genuine solutions. Additionally, those approximations make a contribution to a higher information of the behavior of the studied structures by providing more stable and comprehensive answers. When in comparison to conventional answers, Padé approximations show off advanced performance throughout a number of situations, highlighting the importance of choosing the right numerical approach based on the nature of the hassle. This approach plays a crucial role in scientific and engineering fields that require high precision in modeling and analysis. Overall, the research emphasizes that Padé approximations represent an advanced and reliable option for addressing complex differential equations, opening new avenues for understanding mathematical and physical phenomena more effectively.