Specific Model Research Articles

Effectiveness of Regional Insurance Networks and the European Stability Mechanism

The article is dedicated to approaches to assessing the effectiveness of re- gional insurance networks in comparison with classic financial institutions established under the auspices of the UN – International Monetary Fund and World Bank Group. On the basis of the evolution of the anti-crisis tools of international financial organisations, the place of regional financial mechanisms (regional insurance networks) is analysed as a new form of debasement of financial imbalances of the region, mitigating certain shortcomings of the IMF anti-crisis regulation model. Close attention is paid to the taxonomy of the Regional Financial Arrangements (RFA) in terms of forming a resource base and directions of the use of funds, their integration into global safety networks. The European Stability Mechanism is understood as a specified form of the RFA, which has the largest resources, the possibility of attracting them on a market basis and an extended range of areas of use. The author used specially tailored criteria to assess the effectiveness of the RFA in terms of lock of crisis events at the regional level. It is concluded that the RFA can only partially be considered as a lender of last resort for a group of major states, while a special model of the functioning of the European Stability Mechanism allowed to solve the tasks of anti-crisis regulation in the EU.

Assessment of cloud service trusted state based on fuzzy entropy and Markov chain

In the era of cloud service popularization, the trustworthiness of service is particularly important. If users cannot prevent the potential trustworthiness problem of the service during long-term use, once the trustworthiness problem occurs, it will cause significant losses. In order to objectively assess the cloud service trustworthiness, and predict its change, this paper establishes a special hierarchical model of cloud service trustworthiness attributes. This paper proposes corresponding management countermeasures around the model, defines the cloud service trustworthiness level, defines the cloud service trusted state based on fuzzy entropy and Markov chain, constructs the membership function of the cloud service trusted state, and realizes the assessment of cloud service trustworthiness and its changes according to the prediction method of Markov chain. Through case analysis and method comparison, it shows that the method proposed in this paper is effective and feasible. This method can provide objective and comprehensive assessment data for the cloud service trustworthiness and its change, makes up the deficiency of fuzzy entropy assessment method. This research has important reference value and significance for the research of cloud service trustworthiness assessment.

Evolving Neural Network Designs with Genetic Algorithms: Applications in Image Classification, NLP, and Reinforcement Learning

A method of evolving deep learning architectures using genetic algorithms is presented. The method is a first step towards a low-cost evolutionary search for task-specific neural networks. We evolve task-specific model architectures optimized for fast execution and low error on several standard machine learning tasks: image classification, character-level language modeling, and solving the cart pole problem. We also introduce a simple variation of the method that is capable of evolving neural networks with recurrent connections of varying depth and length and show performance on a word-level language modeling task. The method is implemented in an open-source library. We hope that the ability to run an evolutionary search at this scale will make it possible for a wide audience to develop deep learning architectures that are specialized for a variety of tasks and to develop many interesting novel architectural features. A new method that uses evolutionary search to directly modify existing neural network architectures to perform a specific task is presented. We demonstrate that task-specific specialization of deep learning models can be useful in practice. We modify convolutional neural networks, residual networks, and an LSTM variant to perform various tasks, and show that specialized networks often perform better than models trained from scratch that have many more parameters and much larger training time. For example, on the object recognition task, a specialized model is built by training a base network to predict object position and then applying a series of genetic search operations to squeeze the network and fit new final layer weights to the output. The specialized model is 8 times faster and has 13% lower error, despite being 17 times smaller than a fully trained larger and slower network.

Mechanism of transition from subdiffusive transport to normal diffusion for passive tracers in a plane periodical flow induced by Gaussian noise

The known analytical solution describing a two-dimensional viscous flow with vortices under a driving force is generalized. It is shown that a periodic pattern of asymmetric vortices arises when the force amplitude exceeds a critical value. The transport of an ensemble of passive particles through the resulting structure has been studied. The envelope of each vortex contains a stagnation point, and the tracer motion along it without noise is infinitely long. It is shown that the macroscopic transport of an ensemble of tracers is extremely slow in the noise-free case. The latter leads to the mixing of the tracer ensemble. We study the peculiarities of such transport using a special flow model. Such a model has the form of a series of recurrence mappings. The recurrence mappings are constructed for the ensemble transport through a unit cell using the numerical solution of the Langevin equations in the presence of random particle displacements caused by molecular diffusion. The presence of diffusion allows particles to penetrate into the vortex, which leads to an additional slowing down of transport. The effect of noise on the variance of the tracer ensemble is investigated. It is shown that the growth of the standard deviation with time can be described by a power law. On an intermediate time scale, which grows exponentially long, as the noise strength tends to zero, a subdiffusive behavior is observed. The dependence of the subdiffusion exponent on the molecular diffusivity is obtained.

A Fair Approach to Heart Disease Prediction: Leveraging Machine Learning Model

This study focuses on the tasks of diagnosing and predicting diseases, which are crucial, for accurately classifying and treating them by cardiologists. By utilizing the increasing use of machine learning in the field in pattern recognition from data this research introduces a specialized model that aims to predict cardiovascular diseases. The main objectives of this model are to reduce misdiagnosis rates and minimize fatalities. To achieve these goals the proposed approach combines Logistic Regression with a fairness component. The model is trained using a real world dataset consisting of 70,000 instances obtained from Kaggle. The dataset is split into 70% for training and 30% for testing purposes to evaluate accuracy and fairness metrics at values of Logistic Regression. Through reweighing techniques applied to the model improvements, in both accuracy and fairness are observed. In conclusion this research suggests that machine learning models that prioritize fairness demonstrate performance by achieving an accuracy rate of 72% with a fairness value of 0.009.

The effect of alimony on married women's labor supply and fertility: Evidence from state‐level reforms

AbstractReforms that reduce spousal support after divorce are generally thought to reduce the bargaining power of alimony recipients as well as their incentives to participate in the traditional model of household specialization. Using the U.S. Time Use Survey and exploiting a series of recent reforms in several U.S. states that reduced the rights of eligible spouses, we find that wives surprised by the reforms reacted by increasing their labor supply, but the adjustment mechanism varied by educational group. While college educated wives' labor supply increased at the expense of time spent on housework and childcare, wives with less than a college degree sacrificed personal care and leisure time. Effects for men are generally smaller and not statistically significant. We also find a reduction in the rate of new births following the reforms, with the effect being much more pronounced in the case of women with a university education, suggesting that part of the difference in time use may be attributable to differences in preferences and the cost of children by educational levels. Since children are a marital public good, our results imply a decrease in marriage‐specific investment. The estimated effects are robust to various sensitivity tests and are greater in couples with a high‐income differential and therefore more exposed to changes in alimony laws.

Simulation of Ozone Distribution in an Innovative Drying and Sanitising Cabinet Chamber

Common designs of workwear drying units require not only energy efficiency but also effective disinfection. One possibility of sanitising clothes during drying is to use the ozone generated inside the drying chamber. This process requires precise management of airflow and a uniform distribution of ozone in the chamber. Therefore, optimising the shape of the drying chamber must include not only the correct flow of drying air but also the correct distribution of ozone. This paper addresses the difficult problem of modelling the flow of sanitising ozone in an innovative drying chamber. The innovative shape of the chamber is shown in this article. Due to the low percentage of ozone in the air (up to 10 ppm), CFD simulation models of the usual mixture type are too inaccurate; therefore, special models have to be used. Therefore, this paper presents an experimentally verified methodology to simulate ozone flow in an innovative drying and sanitising cabinet using two methods: Discrete Phase Model (DPM) and Species Transport (ST). The DPM method uses a Euler–Lagrange approach to qualitatively assess the spread of ozone particles, treated with a description of the movement of the particles and not as a continuous gaseous substance. On the other hand, this already allows the verification of ozone concentrations, with an appropriate conversion of the measured quantities. The ANSYS/FLUENT 2023R1 program was used for the simulations after careful selection of the mesh, closing models, boundary conditions, etc. Simulations made it possible to analyse the distribution of ozone in the workspace and assess the effectiveness of the sanitisation process. The results of the simulations were verified on the basis of empirical tests, which showed the correctness of the model and the correct distribution of the sanitising ozone in the entire volume of the drying chamber in the innovative drying–sanitising chamber. The complete simulation of the air and ozone distribution using the presented models allowed for the optimisation of the opening and shapes, which contributed to improving the energy efficiency of the unit and increasing the efficiency of the sanitisation processes, making the described methodology very effective for optimising the chambers of various types of dryers.

Properties and Analysis of the Guard Interval in Infinite Impulse Response–Orthogonal Frequency Division Multiplexing Systems

Recently, an orthogonal frequency division multiplexing (OFDM) technique for the infinite impulse response (IIR) channel (IIR-OFDM) was proposed, which carries the dedicated guard interval to maintain the circular convolution of the received signal and channel coefficients. Therefore, the loop of an IIR channel can be converted to the frequency domain, and single-tap equalization can still be used to equalize loop interference, like classical OFDM. In this paper, we describe how to build the IIR system based on the channel with loops and derive the properties of the dedicated guard interval for a general multi-order IIR channel, which is different from the classical cyclic prefix (CP) obtained by replicating the samples at the tail end of the signal. In particular, we address two special models for first-order and delay IIR channels. It is demonstrated that the guard interval composition and power characteristics of the two special models are similar. Moreover, the complexity of the guard interval depends not only on the maximum delay of the loop, but also on the number of loops. Finally, we simulate the IIR-OFDM performance under different IIR channels.

DIGITAL SUPPORT OF STUDENTS’ JOINT ACTIONS IN INTRODUCTION TO THE CONCEPT’S ORIGIN: THE CASE OF PROPORTIONALITY

Our primary school module with the digital support “Inks” was designed following V.V. Davydov’s idea on the assembling sets context as crucial for number concept formation. Making-sets task and the mixture problem scaffold the initial steps towards proportional reasoning, which is known to be diffi cult for students. The learning situation revolves around purposeful transformations of the inks shade towards the one desired, with a variety of tasks’ conditions: restrictions on the available amount of water and inks, some ready-made mixtures provided, etc. The substantial feature of our instruction was the organization of joint work through distribution of the operations on paint components. The necessity of some preliminary coordination between partners is posed by the real-life restrictions: once water and inks have been mixed, they cannot be separated back. The computer simulation adheres to this “rule” and makes students refer to the special modeling space and plan their operations on water and inks beforehand using tokens. The divergence of results, predicted by students through tokens placement, and the real color shade, obtained in computer simulation, sets the learning task and emphasizes the problem of coordination between partners’ actions within the modeling space. As students start to rely on the portion-by-portion counting, they succeed on the fi rst try: the stain of their new paint, prepared with the different amounts of components, merges with the sample mixture.  second-graders from an urban school (– years old) participated in the experimental teaching series, the video-taped classroom discussions, computer logs and written works were analyzed in order to assess students’ conceptual thinking development. Students’ constant success in tasks and the development of their modeling means functionality proved the computer support, which we have designed, to be relevant in several aspects: problematization and feedback, enabling simultaneous partners’ work, providing substantial tasks variations

Exploring the Application of Large Language Models Based AI Agents in Leakage Detection of Natural Gas Valve Chambers

Leakage problems occur from time to time due to the large number of equipment and complex processes during oil and gas production and transportation. The traditional detection methods highly depend on manpower with large workload and are prone to missed and false alarms, which seriously affects the efficiency and safety of oil and gas production and transportation. With the continuous improvement of information technology and the rapid advancement of artificial intelligence (AI), the research on leakage detection technology based on AI methods has attracted more and more attention. This paper discusses the application scenarios of an AI agent based on the recently emerged large language model (LLM) technology in oil and gas production leakage detection: (1) Compared with the traditional leakage detection methods, this paper innovatively employs a combination of AI-based diagnostics and infrared temperature measurement technologies to develop a specialized small model for oil and gas leakage detection, which has been proven to significantly improve the accuracy of detecting industrial venting events in natural gas valve chambers; (2) By employing retrieval-augmented generation (RAG) technology, a knowledge vector library is constructed, utilizing a series of leakage-related documents, assisting the LLM to carry out knowledge questioning and inference. Compared with the traditional SimBERT, the accuracy can be improved by about 15% in the Q&A search ability test. The correct rate is about 70% higher than the SimBERT in the Chinese complex reasoning quiz. Also, it can still remain stable under high load conditions, with the interruption rate of retrieval closing to zero. (3) By combining the specialized small model and the knowledge Q&A tool, the natural gas valve chambers’ leakage detection AI agent based on the open-source LLM model was designed and developed, which preliminarily achieved the leakage detection based on the specialized small model, and the automatic processing of the retrieval reasoning process based on the knowledge Q&A tool and the intelligent generation of corresponding leakage disposal scheme. The effectiveness of the method has been verified by actual project data. This article conducts preliminary explorations into the in-depth applications of AI agents based on LLMs in the oil and gas energy industry, demonstrating certain positive outcomes.

Modeling Borehole Resistivity Logs with Exponentially Graded Multilayered Soil Inversion for Electrical Engineering Applications

In this paper, we present the application of a special soil resistivity model, the Exponentially Graded Multilayered Soil model, to the interpretation of Borehole Resistivity Logging surveys. The model, which was recently introduced by the authors, offers an alternative to the interpretation provided by constant resistivity layer models. The model is applied to synthetic soil surveys as well as to a real survey that is not satisfactorily interpreted using conventional models. This model aids in the design and calculation of grounding systems in electrical engineering, providing an alternative to calculations performed by traditional layered models.

Investigation of the Items Prepared by Teachers Who Received Training on Item Preparation Measuring Higher-Order Thinking Skills

This study was conducted to determine how the in-service training program for science teachers affects teachers' competencies in preparing items measuring higher order learning levels. The study was conducted with the exploratory special case design, which is one of the special case study models. The study group consisted of 8 science teachers who participated in the training program implemented within the scope of the study. The training program is an 8-week question preparation training that measures higher-order thinking skills. The training was carried out theoretically and practically by science education, assessment and evaluation and curriculum experts. During the study, the teachers prepared items to measure higher order thinking skills related to the Earth and Universe learning area in the science curriculum for six months. A 21-item checklist developed by the researcher was used as a data collection tool in the study. The data obtained in the scope of the program were analyzed descriptively. The analyses were reported as frequency and percentage, and the results were visualised with graphs. At the end of the program, it was determined that 43% of the items prepared by the teachers measured higher-order thinking skills. This finding suggests that the training program positively affected the teachers' competence in writing items that measure higher-order thinking skills. At the end of the study, a fair number of suggestions were made, the most important being the widespread implementation of the training program.

Zebrafish as a Model Organism for Congenital Hydrocephalus: Characteristics and Insights.

Hydrocephalus is a cerebrospinal fluid-related disease that usually manifests as abnormal dilation of the ventricles, with a triad of clinical findings including walking difficulty, reduced attention span, and urinary frequency or incontinence. The onset of congenital hydrocephalus is closely related to mutations in genes that regulate brain development. Currently, our understanding of the mechanisms of congenital hydrocephalus remains limited, and the prognosis of existing treatments is unsatisfactory. Additionally, there are no suitable or dedicated model organisms for congenital hydrocephalus. Therefore, it is significant to determine the mechanism and develop special animal models of congenital hydrocephalus. Recently, zebrafish have emerged as a popular model organism in many fields, including developmental biology, genetics, and toxicology. Its genome shares high similarity with that of humans, and it has fast and low-cost reproduction. These advantages make it suitable for studying the pathogenesis and therapeutic approaches for various diseases, specifically congenital diseases. This study explored the possibility of using zebrafish as a model organism for congenital hydrocephalus. This review describes the characteristics of zebrafish and discusses specific congenital hydrocephalus models. The advantages and limitations of using zebrafish for hydrocephalus research are highlighted, and insights for further model development are provided.

Tuning Electron-Transfer Driving Force in Photosynthetic Special Pair Models.

Visible-light excitation of a family of bimetallic ruthenium polypyridines with the formula [RuII(tpy)(bpy)(𝜇-CN)RuII(py)4L]n+ (RuRuLn+), where L = Cl-, NCS-, DMAP and ACN, was used to prepare photoinduced mixed-valence (PI-MV) MLCT states as models of the photosynthetic reaction center. Ultrafast transient absorption spectroscopy allowed to monitor photoinduced IVCT bands between 6000 and 11000 cm-1. Mulliken spin densities resulting from DFT and (TD)DFT computations revealed the modulation of the charge density distribution depending on the ligand substitution pattern. Results are consistent with PI-MV systems ranging from non-degenerate Class II to degenerate Class III, with electronic couplings between 1000 and 3500 cm-1.These findings guide the control electron localization-delocalization in charge-transfer/charge-separated excited states, like those involved in the photosynthetic reaction center.

Feasibility Testing of Digitise Home Visit Service(DHVS) Core Model

The COVID-19 pandemic precipitated a global surge in homeschooling (Ray et al., 2020). While support networks for homeschooling parents have expanded, access to in-home educational professionals, particularly in rural areas, remains a significant challenge (Bullock & Davis, 2018). This study introduces a novel "Digitized Home Visit Education Services (DHVS) (Sathya D H, 2024)" model to address this gap. Building on two years of experience delivering one- to-one home visit special education, our model integrates online monitoring with in-home support from a local trained resource person. The model's feasibility was assessed through a comprehensive evaluation, encompassing subject selection, learning profile analysis, teacher training, and online monitoring during home visits. Semi-structured interviews with the parent, the teacher, the trained resource person, and the passive observer provide qualitative data to refine the model. This paper delves into the Digitised Home Visit Service (DHVS) Core Model, evaluating its feasibility and acceptability. Findings will contribute to the development of DHVS as an alternative special education and therapy service delivery model (Sathya D H, 2024), informing rehabilitation and special education professionals about an innovative method leveraging technology to reach needy students. Parents of students with disabilities will appreciate the digitised home visit model when it becomes available for their children to receive education at home.

Alleviating H 0 and S 8 Tensions Simultaneously in K-essence Cosmology

The present work begins by examining the early-Universe inflationary epoch of a special K-essence model, which incorporates a linear coupling term between the scalar field potential and the canonical Lagrangian. For the power-law potential, we both numerically and analytically prove that the inflationary parameters such as the spectral index and tensor-to-scalar ratio are compatible with the recent BICEP/Keck observations. Continuing this work, our analysis based on comparing early-Universe observations with late-Universe measurements indicates that the tension on the Hubble parameter H 0 and the growth of structure parameter S 8 can be alleviated simultaneously. More precisely, compared to the standard ΛCDM model, our model can reduce H 0 tension to roughly 2.2σ, and the S 8 discrepancy diminishes to 0.82σ.

3D printing of special shape model: New insight into material properties and correlation of model printing properties

Improving the ability to print complex shapes has been a key focus in the development of 3D food printing. This study evaluated the material properties and printing characteristics of various materials, using the printing data from test models to create a dataset for categorizing, predicting, and validating the special shape printing capabilities of these materials. The primary factors influencing printing performance were the material's mechanical strength, resistance to deformation, and the ‘extrusion expansion’ effect. The print accuracy and support properties of test models of print material can significantly affect its ability to print special shapes. Based on special shape modeling experiments, the printing materials were classified into three categories: gels that could not print either flat or three-dimensional models, gels that could only print flat models, and gels that could print both flat and three-dimensional models. Principal Component Analysis (PCA) and Fisher Discriminant Analysis (with classification accuracies of 98.33% for the training set and 96.67% for the prediction set) confirmed the validity of these classification results. This study established a correlation between the printing characteristics of test models and those of special shape models, aiding in the evaluation of the printability of complex shapes.

A Dirichlet Process Model for Directional-linear Data with Application to Bloodstain Pattern Analysis

Directional data require specialized models because of the non-Euclidean nature of their domain. When a directional variable is observed jointly with linear variables, modeling their dependence adds an additional layer of complexity. A Bayesian nonparametric approach is introduced to analyze directional-linear data. Firstly, the projected normal distribution is extended to model the joint distribution of linear variables and a directional variable with arbitrary dimension projected from a higher-dimensional augmented multivariate normal distribution. The new distribution is called the semi-projected normal distribution (SPN) and can be used as the mixture distribution in a Dirichlet process model to obtain a more flexible class of models for directional-linear data. Then, a conditional inverse-Wishart distribution is proposed as part of the prior distribution to address an identifiability issue inherited from the projected normal and preserve conjugacy with the SPN. The SPN mixture model shows superior performance in clustering on synthetic data compared to the semi-wrapped Gaussian model. The experiments show the ability of the SPN mixture model to characterize bloodstain patterns. A hierarchical Dirichlet process model with the SPN distribution is built to estimate the likelihood of bloodstain patterns under a posited causal mechanism for use in a likelihood ratio approach to the analysis of forensic bloodstain pattern evidence.

ГІБРИДНИЙ МЕТОД АНАЛІЗУ ТА ІДЕНТИФІКАЦІЇ ШКІДЛИВОГО DOH-ТРАФІКУ

This study addresses the challenges associated with detecting DNS over HTTPS (DoH) traffic, a relatively new protocol that has not been extensively researched. The detection methods discussed include TLS inspection, application logging, and open-source tools such as Zeek and RITA. TLS inspection, which involves decrypting and analyzing traffic, is the most intrusive and requires full control over the network and client configurations. Application logging, such as that available in Mozilla Firefox, necessitates administrative control over client systems, which may be impractical. Zeek analyzes network logs to identify domains accessed without regular DNS queries, while JA3 fingerprints and RITA focus on detecting malicious DoH traffic by analyzing TLS handshake parameters and beacon-like activities, respectively. Additionally, maintaining up-to-date blacklists of IP addresses and SNI values can help identify DoH traffic but faces scalability and evasion challenges. The study highlights that no current solution is entirely feasible, with many requiring excessive administrative overhead or failing to scale effectively. A hybrid approach using machine learning models and traffic analysis, as illustrated by the CIRA-CIC-DoHBrw-2020 dataset, is proposed for more effective detection of malicious DoH traffic. This approach involves the architecture of a two-stage DoH traffic identification system is presented, consisting of three subsystems: traffic, training and evaluation, and identification. They operate sequentially, with the system's function being traffic identification, training, testing, and information processing within the DoH protocol. The next step is process of cross-validation, which involves training a machine learning model K times, with each iteration using a different fold as the validation set, while the remaining folds serve as the training set. The aim of this work: Development and implementation a DoH traffic identification system, which, unlike existing solutions, is based on a hybrid approach to identifying malicious traffic using open tools for detecting encrypted DNS traffic and specialized machine learning models.

A Multi-model Deep Learning Architecture for Diagnosing Multi-class Skin Diseases.

Skin diseases are a significant global public health concern, affecting 21-85% of the world's population, particularly those in low- and middle-income countries. Accurate and timely diagnosis is crucial for effective treatment and improved patient outcomes. This study introduces a novel deep-learning multi-model architecture designed for high-precision skin disease diagnosis. The system employs a five-category Xception model to classify skin lesions into five classes: Atopic Dermatitis, Acne and Rosacea, Skin Cancer, Bullous, and Others. Trained on 25,010 images, the model achieved 95% accuracy and an AUROC of 99.4%. To further enhance accuracy, transfer learning was applied, resulting in specialized models for each class, with strong performance across 40 skin conditions. Specifically, the Acne and Rosacea model achieved an accuracy of 90.0%, with a precision of 90.7%, recall of 90.1%, f1-score of 90.2%, and an AUROC of 99.0%. The Skin Cancer model demonstrated 94.0% accuracy, 94.8% precision, 94.2% recall, 94.1% f1-score, and a 99.5% AUROC. The Atopic Dermatitis model reported 91.8% accuracy, 92.2% precision, 91.8% recall, 91.9% f1-score, and a 98.8% AUROC. Finally, the Bullous model showed 90.0% accuracy, 90.6% precision, 90.0% recall, 90.0% f1-score, and a 98.9% AUROC. This approach surpasses previous studies, offering a more comprehensive diagnostic tool for skin diseases. To facilitate result reproducibility, the training and testing codes for the models utilized in this study are accessible via the GitHub repository ( https://github.com/SaraEl-Metwally/A-Multi-Model-Deep-Learning-for-Diagnosing-Skin-Diseases ).