Scientific Domains Research Articles

Beyond The Basics: A Detailed Survey of Advanced Python Applications and Innovations

Abstract: Python is the general purpose, high-level, scriptable, interpreted, object oriented and easy to learn language was developed by Guido van Rossum. Python is a popular programming language because of its simplicity, ease of use, open-source licensing, and accessibility. It provides various libraries and frameworks for learning and real world programming. Python’s versatility and active community make it an ideal language for Machine Learning, Data Science, Artificial Intelligence, Networking, Game Development, Web Development and various other domains. In this paper, we will provide an introduction to the main Python programming software tools used for Web development, Data science, Machine learning techniques, Educational Development, etc. In this paper will first introduce Python as a language, and give introduction about Web Development, Data science, Machine learning, Educational Development & Financial Analysis and then describe packages that are popular in that sector

A Comparative Study on ERNIE Bot 4.0 Turbo and ChatGPT 4o’s Performance in Evaluating First-Year Undergraduate Persuasive Essays

This study compares the performance of two prominent AI language models, ERNIE Bot 4.0 Turbo and ChatGPT 4o, in evaluating first-year undergraduate persuasive essays within the social sciences domain. Drawing from the Louvain Corpus of Native English Essays, a comprehensive collection of academic writings by British and American university students, this study aims to examine the models’ capabilities in assessing the grammatical correctness, vocabulary usage, coherence, content depth, and writing style of the essays. This study adopts a structured evaluation framework based on IELTS writing criteria to assess the models’ performance. A 40 persuasive essays from the Louvain Corpus were evaluated by both AI models and compared with human raters’ evaluations to ensure validity. The findings reveal distinct differences in the assessment styles of the two models. ChatGPT 4o exhibits a more critical approach, pinpointing areas for improvement, such as lack of argument development, coherence issues, and grammatical errors. Conversely, ERNIE Bot 4.0 Turbo offers a more balanced assessment, acknowledging essays’ strengths and suggesting improvement areas. Notably, ERNIE Bot’s evaluation highlights potential biases in AI-based assessment systems, particularly in its unequal emphasis on viewpoints. This comparative examination offers valuable perspectives on the advantages and constraints of AI models in assessing scholarly compositions, underscoring the significance of amalgamating varied AI functionalities to establish more all-inclusive and efficient feedback systems for learners. By understanding these differences, researchers and educators can better utilize AI-assisted essay evaluation systems to enhance student learning experiences.

On the Use of Image Analysis for Hematocrit Evaluation in Dried Blood Spots

Dried blood spots (DBSs) are formed by collecting a small sample of blood on specialized filter paper and allowing it to dry naturally. Various domains of life sciences and drug research extensively use DBSs as a sampling technique. The “Hematocrit (Ht) effect” affects assay bias, and several strategies have been put forth to deal with it, including the correction of quantified concentrations using an appropriate correction factor. The approach was previously applied, following the utilization of an image processing algorithm developed in Matlab® to derive a reliable equation correlating DBS areas to Ht% values. The present work looks more closely at the application of image analysis to the evaluation of Ht in DBS samples. Utilizing image analysis software, DBS samples with known Ht values were processed. Preparation of cards has followed a previously developed protocol for the appropriate formation of uniform area DBSs, irrespective of Ht. The resulting areas showed close resemblance to the respective theoretical areas calculated by applying the correlation equation. Following that, the equation was utilized to determine the Ht values for each sample, and a comprehensive comparison of measured versus calculated Ht was carried out using various statistical approaches for method comparison. The results demonstrated a strong correlation, suggesting the method’s viability in estimating Ht for unknown DBS samples.

A vulnerable soil environment study in karst areas: a bibliometric analysis

Karst landforms are widely distributed around the world, and karst rocky desertification has occurred on a large scale in many countries and regions, causing significant adverse impacts on local natural environments and societies. The improvement and rational use of karst soil is a key aspect of rocky desertification governance. Karst soil science studies are of great value in karst regions and are essential for controlling karst rocky desertification and ecological restoration. In order to understand the research hotspots and the development directions in the field of vulnerable karst soil environment, we undertook bibliometrics citation analysis on 1913 contributions to the literature written in the range from 2001 to 2019 based on the “Web of Science” core collection citation index database. Hopefully, this work will help to set up a scientific foundation for further studies. Using CiteSpace visualization software, we analyzed the distribution of disciplinary categories, reference co-citation clusters, and keyword clusters in the literature. The results show the basic characteristics and evolution of the literature related to karst pedology. We then recognized the main intellectual bases in the domain of karst soil science. This study also revealed the research hotspots and trends in this field. Through a bibliometrics citation analysis of research on karst vulnerable soil environment, the present study provides a quantitative and objective understanding of development directions that have emerged in this field over the past 19 years, offering a reference for future research.

9.X.2. Activating the ASPHER-WHO Public Health Core Curriculum: Interprofessional Skills

Abstract The updated ASPHER-WHO Public Health Core Curriculum builds on ASPHER’s previous work on Competencies for Public Health Professionals (2006-2020), reflects the changing needs of our population and planet and public health professionals. The key objectives of this workshop are to: - Share the updated ASPHER core curriculum for public health; - Illustrate examples of public health challenges that young professionals have championed for inclusion; - Share the Faculty of Public Health (UK)’s work on Fair training as a case study on operationalising public health leadership on equality, diversity and inclusion and demonstrating the core competencies in action. A curricular content survey of member schools and programmes yielded responses from 60 schools, with > 500 submissions of detailed. subject curricula. These have been compiled into 36 Subject Areas in four Domains, namely: (1) Core Subject Areas in Public Health: (2) Subject-specific Areas in Public Health;(3) Core Cross-curricular Subject Areas; and (4) Core Interdisciplinary Professional Skills in PH. The development process was guided and grounded by the expertise and passion of young public health professionals. They are helping us make the core curriculum fit for purpose for the near future and provide them with the skills required to be a competent member of a supportive public health team in an ever changing world. This Workshop will profile selected Subject Areas from each Domain using real life challenges facing WHO European Region to show how the core curriculum integrates research, theory and practice to help build those essential interdisciplinary skills. The examples will focus on Economics in Public Health (Domain 1); Public Health and Criminal Justice; (Domain 2); Health Literacy and Infodemiology, Mental Health and Wellbeing (Domain 3); and Integrative Learning in Action; Communication Skills; Advocacy, Negotiation; Collaborations and Partnerships; and Leadership, Management and Implementation Science (Domain 4). Discussions for each of the Subject Areas will be led by an expert in the area. We will also illustrate an example of these competencies in action, how they are helping shine a light on the challenges we face in creating that supportive team and environment and how the potential for collaboration helps us to capitalise on the expertise available in some schools and programmes for the benefit of all. Key messages • Learning interdisciplinary skills are essential for effective public health analysis, advocacy and action. • Collaboration across countries, schools, professions and generations can create transformational change. Speakers/Panelists Mary Codd University College Dublin, Dublin, Ireland Karl F Conyard University College Dublin, Dublin, Ireland Jenny Houston University of Edinburgh, Edinburgh, UK Samia Latif Faculty of Public Health, London, UK Olalekan Popoola ASPHER, UCD, Dublin, Ireland

Experimental investigation on the electrical and hydrogen-production performance of a novel PEM electrolyzer driven by CdTe PV with a tracking system

Addressing the challenges of intermittent and variable energy supply, solar hydrogen production via electrolysis has garnered significant attention due to its unique advantages, emerging as a hot topic in the scientific research domain. However, the application of this technology is currently mainly related to the non-tracking system, which undergoes a rapid change in solar irradiation during the day. To expand the matching relationship of solar irradiation, PV characteristics, and electrolyzer, this paper proposes the development of a Proton Exchange Membrane (PEM) electrolyzer driven by Cadmium Telluride (CdTe) photovoltaic (PV) modules with a tracking system for hydrogen production. An experimental platform was set up in Chengdu and the performance tests were carried out for different operating modes during summer. The experimental results demonstrate that the average electrical and hydrogen production efficiencies of the system are 13.11% and 7.88%, respectively, on a sunny day, with a total of 566.7 L of hydrogen produced within 7 h of operation. Furthermore, the solar irradiation received by the tracking surface of the system is significantly enhanced by 10.29% in comparison to the horizontal surface. On a cloudy day with an electrolyzer inlet water temperature of 45 °C, the system has the potential to achieve a hydrogen production efficiency of 11.15 %. At a constant current density of 1.0 A/cm2, the efficiency of the electrolysis process increased by 2.07% and 1.25% for each 10 °C increase in the temperature of the electrolyzer, having an inlet temperature of 35 °C. This study offers a novel approach for a PV-PEM system with a higher hydrogen production efficiency while also establishing a robust foundation for technology optimization and broader implementation.

AS2LS: Adaptive Anatomical Structure-based Two-layer Level Set Framework for Medical Image Segmentation.

Medical images often exhibit intricate structures, inhomogeneous intensity, significant noise and blurred edges, presenting challenges for medical image segmentation. Several segmentation algorithms grounded in mathematics, computer science, and medical domains have been proposed to address this matter; nevertheless, there is still considerable scope for improvement. This paper proposes a novel adaptive anatomical structure-based two-layer level set framework (AS2LS) for segmenting organs with concentric structures, such as the left ventricle and the fundus. By adaptive fitting region and edge intensity information, the AS2LS achieves high accuracy in segmenting complex medical images characterized by inhomogeneous intensity, blurred boundaries and interference from surrounding organs. Moreover, we introduce a novel two-layer level set representation based on anatomical structures, coupled with a two-stage level set evolution algorithm. Experimental results demonstrate the superior accuracy of AS2LS in comparison to representative level set methods and deep learning methods.

Exploration of Chemical Space through Automated Reasoning

The vast size of composition space poses a significant challenge for materials chemistry: exhaustive enumeration of potentially interesting compositions is typically infeasible, hindering assessment of important criteria ranging from novelty and stability to cost and performance. We report a tool, Comgen, for the efficient exploration of composition space, which makes use of logical methods from computer science used for proving theorems. We demonstrate how these techniques, which have not previously been applied to materials discovery, can enable reasoning about scientific domain knowledge provided by human experts. Comgen accepts a variety of user‐specified criteria, converts these into an abstract form, and utilises a powerful automated reasoning algorithm to identify compositions that satisfy these user requirements, or prove that the requirements cannot be simultaneously satisfied. In contrast to machine learning techniques, explicitly reasoning about domain knowledge, rather than making inferences from data, ensures that Comgen’s outputs are fully interpretable and provably correct. Users interact with Comgen through a high‐level Python interface. We illustrate use of the tool with several case studies focused on the search for new ionic conductors. Further, we demonstrate the integration of Comgen into an end‐to‐end automated workflow to propose and evaluate candidate compositions quantitatively, prior to experimental investigation.

Exploration of Chemical Space through Automated Reasoning.

The vast size of composition space poses a significant challenge for materials chemistry: exhaustive enumeration of potentially interesting compositions is typically infeasible, hindering assessment of important criteria ranging from novelty and stability to cost and performance. We report a tool, Comgen, for the efficient exploration of composition space, which makes use of logical methods from computer science used for proving theorems. We demonstrate how these techniques, which have not previously been applied to materials discovery, can enable reasoning about scientific domain knowledge provided by human experts. Comgen accepts a variety of user-specified criteria, converts these into an abstract form, and utilises a powerful automated reasoning algorithm to identify compositions that satisfy these user requirements, or prove that the requirements cannot be simultaneously satisfied. In contrast to machine learning techniques, explicitly reasoning about domain knowledge, rather than making inferences from data, ensures that Comgen's outputs are fully interpretable and provably correct. Users interact with Comgen through a high-level Python interface. We illustrate use of the tool with several case studies focused on the search for new ionic conductors. Further, we demonstrate the integration of Comgen into an end-to-end automated workflow to propose and evaluate candidate compositions quantitatively, prior to experimental investigation.

Unveiling the Potential of Protein-Based Sustainable Antibacterial Materials.

The surge in bacterial growth and the escalating resistance against a multitude of antibiotic drugs have burgeoned into an alarming global threat, necessitating urgent and innovative interventions. In response to this peril, scientists have embarked on the development of advanced biocompatible antibacterial materials, aiming to counteract not only bacterial infections but also the pervasive issue of food spoilage resulting from microbial proliferation. Protein-based biopolymers and their meticulously engineered composites are at the forefront of this endeavor. Their potential in combating this severe global concern presents an approach that intersects the domains of biomedicine and environmental science. The present review article delves into the intricate extraction processes employed to derive various proteins from their natural sources, unraveling the complex biochemical pathways that underpin their antibacterial properties. Expanding on the foundational knowledge, the review also provides a comprehensive synthesis of functionalized proteins modified to enhance their antibacterial efficacy, unveiling a realm of possibilities for tailoring solutions to specific biomedical and environmental applications. The present review navigates through their antibacterial applications; from wound dressings to packaging materials with inherent antibacterial properties, the potential applications underscore the versatility and adaptability of these materials. Moreover, this comprehensive review serves as a valuable roadmap, guiding future research endeavors in reshaping the landscape of natural antibacterial materials on a global scale.

Analysis of multi-term arbitrary order implicit differential equations with variable type delay

Due to their capacity to simulate intricate dynamic systems containing memory effects and non-local interactions, fractional differential equations have attracted a great deal of attention lately. This study examines multi-term fractional differential equations with variable type delay with the goal of illuminating their complex dynamics and analytical characteristics. The introduction to fractional calculus and the justification for its use in many scientific and technical domains sets the stage for the remainder of the essay. It describes the importance of including variable type delay in differential equations and then applying it to model more sophisticated and realistic behaviours of real-world phenomena. The research study then presents the mathematical formulation of variable type delay and multi-term fractional differential equations. The system’s novelty stems from its unique combination of variable delay, generalized multi terms fractional differential operators (n and m), and integral implicit parameters, and studying the stability of the the newly formulated system as compared to the work in the existing literature. While the variable type delay is introduced as a function of time to describe instances where the delay is not constant, the fractional order derivatives are generated using the Caputo approach. The existence, uniqueness, and stability of solutions are the main topics of the theoretical analysis of the suggested differential equations. In order to establish important mathematical features, the inquiry makes use of spectral techniques, and fixed-point theorems. The study finishes by summarizing the major discoveries and outlining potential future research avenues in this developing field. It highlights the potential contribution of multi-term fractional differential equations with variable type delay to improving the control and design of complex systems. Overall, this study adds to the growing body of knowledge in the field of fractional calculus and provides insightful information about the investigation of multi-term fractional differential equations with variable type delay, making it pertinent for academics and practitioners from a variety of fields.

Developmental and validation of a novel small and high-efficient panel of microhaplotypes for forensic genetics by the next generation sequencing

BackgroundIn the domain of forensic science, the application of kinship identification and mixture deconvolution techniques are of critical importance, providing robust scientific evidence for the resolution of complex cases. Microhaplotypes, as the emerging class of genetic markers, have been widely studied in forensics due to their high polymorphisms and excellent stability.Results and discussionIn this research, a novel and high-efficient panel integrating 33 microhaplotype loci along with a sex-determining locus was developed by the next generation sequencing technology. In addition, we also assessed its forensic utility and delved into its capacity for kinship analysis and mixture deconvolution. The average effective number of alleles (Ae) of the 33 microhaplotype loci in the Guizhou Han population was 6.06, and the Ae values of 30 loci were greater than 5. The cumulative power of discrimination and cumulative power of exclusion values of the novel panel in the Guizhou Han population were 1-5.6 × 10− 43 and 1-1.6 × 10− 15, respectively. In the simulated kinship analysis, the panel could effectively distinguish between parent-child, full-sibling, half-sibling, grandfather-grandson, aunt-nephew and unrelated individuals, but uncertainty rates clearly increased when distinguishing between first cousins and unrelated individuals. For the mixtures, the novel panel had demonstrated excellent performance in estimating the number of contributors of mixtures with 1 to 5 contributors in combination with the machine learning methods.ConclusionsIn summary, we have developed a small and high-efficient panel for forensic genetics, which could provide novel insights into forensic complex kinships testing and mixture deconvolution.

Two-Dimensional Hydration and Triple-Interlayer Lattice Structures in Sulfate-Intercalated Graphene Oxide Nanosheets

Sulfate anions (SO42−) are pivotal in various scientific and industrial domains, including mineralogy, biology, and materials science. While extensive research has elucidated sulfate hydration in bulk solids, liquids, and gaseous clusters, a significant gap persists in understanding sulfate interactions within two-dimensional materials, particularly graphene oxide (GO) nanosheets. This study investigates the intricate hydration phenomena and novel triple-interlayer lattice configurations that emerge from sulfate intercalation in GO nanosheets. Utilizing a straightforward methodology for obtaining precise X-ray measurements of confined nanospaces, we analyzed the temperature-dependent behavior and structural characteristics of these systems. Our findings reveal how sulfate ions modulate interlayer spacing, the dynamics of GO layers, and phase transitions. This research offers an atomic-scale understanding of hybrid hydration behaviors within confined SO4-H2O nano-environments, advancing our knowledge of sulfate interactions in two-dimensional materials.

From greatest simplicity to full power

AbstractWhile language processing services are key assets for the science and technology of language, the possible ways under which they may be made available to the widest range of their end users are critical to support an Open Science policy for this scientific domain. Although providing such processing services under some web-based interface, at large, offers itself as an immediate and cogent response to that challenge, turning this view into an effective access to language processing services is an undertaking deserving a clear conceptual direction and a corresponding robust empirical validation. Based on an extensive overview of major undertakings towards making language processing tools available and on the design principles worked out and implemented in the PORTULAN CLARIN infrastructure, in this paper we advocate for a Research-Infrastructure-as-a-Service (RIaaS) model. This model unleashes accessibility to language processing services in as many web-based interface modalities as the current stage of technological development permits to support, in order to serve as many types of end users as possible, from IT developers to Digital Humanities researchers, and including citizen scientists, teachers, students and digital artists among many others.

Exploring the Need for Medical Futures Studies: Insights From a Scoping Review of Health Care Foresight.

The historical development and contemporary instances of futures studies, an interdisciplinary field that focuses on exploring and formulating alternative futures, exemplify the increasing significance of using futures methods in shaping the health care domain. Despite the wide array of these methodologies, there have been limited endeavors to employ them within the medical community thus far. We undertook the first scoping review to date about the application of futures methodologies and published foresight projects in health care. Through the use of the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) method, we identified 59 studies that were subsequently categorized into the following 5 distinct themes: national strategies (n=19), strategic health care foresight (n=15), health care policy and workforce dynamics (n=6), pandemic preparedness and response (n=7), and specialized medical domains (n=12). Our scoping review revealed that the application of futures methods and foresight has been successfully demonstrated in a wide range of fields, including national strategies, policy formulation, global threat preparedness, and technological advancements. The results of our review indicate that a total of 8 futures methods have already been used in medicine and health care, while there are more than 50 futures methods available. It may underscore the notion that the field is unexploited. Furthermore, the absence of structured methodologies and principles for employing foresight and futures techniques in the health care domain warrants the creation of medical futures studies as a separate scientific subfield within the broad domains of health care, medicine, and life sciences. This subfield would focus on the analysis of emerging technological trends, the evaluation of policy implications, and the proactive anticipation and mitigation of potential challenges. Futures studies can significantly enhance medical science by addressing a crucial deficiency in the promotion of democratic participation, facilitating interdisciplinary dialogue, and shaping alternative futures. To further contribute to the development of a new research community in medical futures studies, it is recommended to establish a specialized scientific journal. Additionally, appointing dedicated futurists in decision-making and national strategy, and incorporating futures methods into the medical curriculum could be beneficial.

Optimising nanoparticles: harnessing quality by design through 3-level factorial design and box-behnken methodology

Nanoparticles (NPs) have become essential elements in a number of scientific and industrial domains, such as materials research, drug delivery, and diagnostics, because of their special qualities and potential for focused uses. An ideal formulation procedure is essential to maximising the effectiveness and efficiency of nanoparticles. In order to optimise the formulation and production processes for nanoparticles, this research investigates the application of Quality by Design (QbD) principles in conjunction with sophisticated statistical approaches, namely the 3-level factorial design and the Box-Behnken methodology. The Quality by Design methodology is a methodical approach that prioritises predetermined goals, in-depth process comprehension, and careful control grounded in reliable science and quality risk management. When QbD is incorporated into nanoparticle optimisation, reliable and repeatable formulations that adhere to specifications with little variation are guaranteed. To look into how different factors interact and affect the properties of nanoparticles, the 3-level factorial design is used. With this design approach, a thorough understanding of the process variables can be obtained by exploring a broad variety of component values. Critical quality attributes (CQAs) include important parameters such drug encapsulation efficiency, zeta potential, polydispersity index (PDI), and particle size. These parameters are methodically examined to determine the ideal amounts of each. To further refine the 3-level factorial design, the Box-Behnken methodology is applied. Without the requirement for extensive combination testing, this response surface methodology (RSM) is especially useful for building second-order polynomial models and investigating quadratic response surfaces. By enabling a more accurate comprehension of the interactions between inputs and responses, the Box-Behnken design improves the optimisation process and makes it easier to identify the ideal circumstances for nanoparticle synthesis. By utilising both approaches in tandem, this work methodically determines ideal formulation parameters that minimise nanoparticle size and polydispersity while optimising stability and drug loading efficiency. The outcomes highlight how important a QbD framework is for directing the creation of reliable nanoparticle systems that function consistently and predictably.

EURAD state-of-the-art report: development and improvement of numerical methods and tools for modeling coupled processes in the field of nuclear waste disposal

The Strategic Research Agenda (SRA; https://www.ejp-eurad.eu/publications/eurad-sra) of the European Joint Programme on Radioactive Waste Management (EURAD; https://www.ejp-eurad.eu/) describes the scientific and technical domains and sub-domains and knowledge management needs of common interest between EURAD participant organizations. Theme number 7 is entitled “Performance assessment, safety case development and safety analyses.” A list of research and development priorities and activities of common interest to be addressed within EURAD for theme 7 have been established. Amongst others, the Understanding and modelling of multi-physical Thermo-Hydro-Mechanical-Chemical coupled processes (THMC) occurring in radioactive waste disposal is a major and permanent issue to support optimization of design and safety case abstraction. To tackle this challenge a research work package entitled “DONUT: Development and improvement of numerical methods and tools for modelling coupled processes” has been conducted within the EURAD join programming initiative. The purpose of this work package is to improve/develop methods or numerical tools in order to go a step further in development of (i) relevant, performant and cutting-edge numerical methods that can easily be implemented in existing or new tools, in order to carry out high-performance computing to facilitate the study of highly coupled processes in large systems, (ii) numerical scale transition schemes for coupled processes, (iii) innovative numerical methods to carry out uncertainty and sensitivity analyses. In this paper the work carried out within the DONUT work package is put in perspective regarding the existing concept and literature on the field. It does not pretend to be exhaustive but rather to put emphasis on particular issues tackled during the project.

Mathematical Information Retrieval: A Review

Mathematical formulas are commonly used to demonstrate theories and basic fundamentals in the Science, Technology, Engineering, and Mathematics (STEM) domain. The burgeoning research in the STEM domain results in the mass production of scientific documents that contain both textual and mathematical terms. In scientific information, the definition of mathematical formulas is expressed through context and symbolic structure that adheres to strong domain-specific notions. Whereas the retrieval of textual information is well-researched, and numerous text-based search engines are present. However, textual information retrieval systems are inadequate for searching scientific information containing mathematical formulas, including simple symbols to complicated mathematical structures. The retrieval of mathematical information is infancy, and it requires the inclusion of new technologies and tools to promote the retrieval of scientific information and the management of digital libraries. This paper provides a comprehensive study of mathematical information retrieval, highlights their challenges and future opportunities.

Mathematical Analysis of a Navier–Stokes Model with a Mittag–Leffler Kernel

In this paper, we establish the existence and uniqueness results of the fractional Navier–Stokes (N-S) evolution equation using the Banach fixed-point theorem, where the fractional order β is in the form of the Atangana–Baleanu–Caputo fractional order. The iterative method combined with the Laplace transform and Sumudu transform is employed to find the exact and approximate solutions of the fractional Navier–Stokes equation of a one-dimensional problem of unsteady flow of a viscous fluid in a tube. In the domains of science and engineering, these methods work well for solving a wide range of linear and nonlinear fractional partial differential equations and provide numerical solutions in terms of power series, with terms that are simple to compute and that quickly converge to the exact solution. After obtaining the solutions using these methods, we use Mathematica software Version 13.0.1.0 to present them graphically. We create two- and three-dimensional plots of the obtained solutions at various values of β and manipulate other variables to visualize and model relationships between the variables. We observe that as the fractional order β becomes closer to the integer order 1, the solutions approach the exact solution. Lastly, we plot a 2D graph of the first-, second-, third-, and fourth-term approximations of the series solution and observe from the graph that as the number of iterations increases, the approximate solutions become close to the series solution of the fourth-term approximation.

Imbalanced Data Parameter Optimization of Convolutional Neural Networks Based on Analysis of Variance

Classifying imbalanced data is important due to the significant practical value of accurately categorizing minority class samples, garnering considerable interest in many scientific domains. This study primarily uses analysis of variance (ANOVA) to investigate the main and interaction effects of different parameters on imbalanced data, aiming to optimize convolutional neural network (CNN) parameters to improve minority class sample recognition. The CIFAR-10 and Fashion-MNIST datasets are used to extract samples with imbalance ratios of 25:1, 15:1, and 1:1. To thoroughly assess model performance on imbalanced data, we employ various evaluation metrics, such as accuracy, recall, F1 score, P-mean, and G-mean. In highly imbalanced datasets, optimizing the learning rate significantly affects all performance metrics. The interaction between the learning rate and kernel size significantly impacts minority class samples in moderately imbalanced datasets. Through parameter optimization, the accuracy of the CNN model on the 25:1 highly imbalanced CIFAR-10 and Fashion-MNIST datasets improves by 14.20% and 5.19% compared to the default model and by 8.21% and 3.87% compared to the undersampling model, respectively, while also enhancing other evaluation metrics for minority classes.