Simple Model Research Articles

Comprehensive Performance Evaluation of Machine Learning Algorithms for detecting DDoS attacks in SDN

<div align="center">Software-Defined Networking (SDN) revolutionizes networking by separating control logic and data forwarding, enhancing security against threats like Distributed Denial of Service (DDoS) attacks. These attacks flood control plane bandwidth, causing SDN network failures. Recent studies emphasize the efficacy of machine learning and statistical approaches in identifying and mitigating these security risks. However, there has been a lack of focus on employing ensembling techniques, amalgamating diverse machine learning models, selecting pertinent features, and utilizing oversampling techniques to balance categorical data. Our study evaluates 20 machine-learning models, emphasizing feature engineering and addressing class imbalance using Synthetic Minority Oversampling TEchnique (SMOTE). The results indicate that Ensemble methods such as LGBM Classifier, Random Forest Classifier, XGB Classifier, Decision Tree Classifier attained near-perfect scores (almost 100%) across all metrics, suggesting potential overfitting. Conversely, models like AdaBoost Classifier, K-Neighbors Classifier, and SVC exhibited slightly lower (99%) but realistic performance, underscoring the intricacy of accurate prediction in cybersecurity. Simpler models, including Logistic Regression, Linear Discriminant Analysis, and Gaussian Naive Bayes, demonstrated moderate to low accuracy, approximately around 70%. These findings stress the imperative need for a nuanced approach in the selection and fine-tuning of machine learning models to ensure effective DDoS detection in SDN environments.</div>

Comparing Moving Averages And Polynomial Regression In Financial Trend Analysis

Trend analysis forms the backbone of studies related to financial markets as they give investors and market participants insights into expected market behavior, aiding in investment decisions. This paper compares two methods for determining financial trends: Moving Averages (MAs) and Polynomial Regression (PR). Widely used technical indicators like Simple Moving Averages (SMA) and Exponential Moving Averages (EMA) are popular due to their ease of understanding and effectiveness in smoothing price data by removing noise. Although simpler models like linear or logistic regression can be used if the data is suitable, polynomial regression is considered more flexible as it fits a non-linear curve to the data. This study aims to examine the effectiveness of these two approaches in detecting and predicting financial trends accurately. The research involves calculating SMA and EMA on historical stock price data and applying polynomial regressions of various degrees. A comprehensive comparison will focus on the fit, precision, and predictability of each method in financial trend analysis

Relationships among village spatial form, electricity use, and photovoltaic potential: Evidence from 120 villages in Huaiyuan County, China

Village areas cover 94.7% of the land in China and are carriers for energy use and renewables. To support national decarbonization strategy, it is essential to devise proper village decarbonization measures. Understanding the relationships among village spatial form, electricity use and photovoltaic potential is critical to provide empirical references. However, due to the scarcity of spatial form and electricity use data, previous studies have been mostly conducted by simplified simulation models, making it difficult to reveal actual relationships accurately. This study proposed a framework to explore the relationship among spatial form, socioeconomic factors, and comprehensive energy use and photovoltaic potential. Electricity use, socioeconomic status, and 3D models of 120 villages in Huaiyuan, China, were collected by investigation and oblique photography. These villages were clustered into three types by the k-means. Photovoltaic potential was analyzed by numerical models upon 3D spatial forms. Quantitative analysis of spatial form, electricity use intensity, electricity generation intensity, net energy use intensity (NEUI) in each cluster were carried out. The average electricity use intensity is 1.87 GWh/m2/y, and the highest energy generation intensity reaches 9.67 GWh/m2/y. Cluster A showed the best performance in NEUI and had the greatest potential for photovoltaic deployment. Average building height, registered resident population, and average building width were the relevant factors to the NEUI in Clusters A, B, and C. The findings provided a framework for other regions to explore factors related to energy use based on actual data, and a reference for low-carbon village policy in China.

MINN: Learning the Dynamics of Differential-Algebraic Equations and Application to Battery Modeling.

The concept of integrating physics-based and data-driven approaches has become popular for modeling sustainable energy systems. However, the existing literature mainly focuses on the data-driven surrogates generated to replace physics-based models. These models often trade accuracy for speed but lack the generalizability, adaptability, and interpretability inherent in physics-based models, which are often indispensable in modeling real-world dynamic systems for optimization and control purposes. We propose a novel machine learning architecture, termed model-integrated neural networks (MINN), that can learn the physics-based dynamics of general autonomous or non-autonomous systems consisting of partial differential-algebraic equations (PDAEs). The obtained architecture systematically solves an unsettled research problem in control-oriented modeling, i.e., how to obtain optimally simplified models that are physically insightful, numerically accurate, and computationally tractable simultaneously. We apply the proposed neural network architecture to model the electrochemical dynamics of lithium-ion batteries and show that MINN is extremely data-efficient to train while being sufficiently generalizable to previously unseen input data, owing to its underlying physical invariants. The MINN battery model has an accuracy comparable to the first principle-based model in predicting both the system outputs and any locally distributed electrochemical behaviors but achieves two orders of magnitude reduction in the solution time.

Heterogenization of a Sandwich [(PW9O34)2Co4(H2O)2]10− in PCN‐222/PCN‐222(M): Exploring the Electron Transfer for Electrocatalytic CO2 Reduction

AbstractIn this study, we designed and prepared polyoxometalate@metal‐organic framework (POM@MOF) composite catalysts through the anchoring of a sandwich POM [(PW9O34)2Co4(H2O)2]10− (shortened as P2W18Co4) to the hexagonal channel of the PCN‐222 (metal‐free) or PCN‐222(M) (M=Fe, Co) frameworks. The composite materials were applied to the electrocatalytic reduction of CO2 reaction (CO2RR) to analyse the effect of incorporating P2W18Co4 on catalytic activity. The P2W18Co4@PCN‐222 composite exhibited enhanced activity across a wide potential range (−0.60~−0.85 V vs. RHE) and an optimal FECO of 72 % at −0.75 V vs. RHE, which was more than double that of PCN‐222 (FECO=33 %). The current density surpassed that of the PCN‐222 precursors by over sixteen times at the same potential. In contrast, the P2W18Co4@PCN‐222(M) composite demonstrated decreased current density, minimal enhancement in CO2RR activity, and a competing HER behaviour. Density functional theory calculations were conducted on simplified models of P2W18Co4@H2‐TCPP and P2W18Co4@M‐TCPP to elucidate the divergent catalytic performances. The findings revealed that while both configurations exhibited the same rate‐limiting step (formation of the *COOH intermediate), a significantly reduced reaction barrier was only observed in the P2W18Co4@H2‐TCPP setup, explaining its substantial activity improvement.

Seismic performance assessment of structural systems in the aftermath of the 2023 Kahramanmaraş earthquakes: Observations and fragility analyses

This article evaluates how different reinforced concrete (RC) building systems in Türkiye behaved during the extreme 2023 Kahramanmaraş earthquakes. The analysis relies on a comprehensive field survey covering 242 RC buildings across various heavily affected locations. Most surveyed buildings were low- and mid-rise RC moment frames and frame-wall (hybrid) systems, with RC wall construction being less commonly observed. Both RC frame and hybrid buildings exhibited several common deficiencies, resulting in significant structural and non-structural damage due to high drift demands. The performance of RC wall construction varied, with some buildings sustaining severe damage while others remained largely unaffected. An analysis of structural plans revealed that RC wall buildings with adequate wall amounts demonstrated exceptional performance, while those with inadequate amounts of walls experienced severe damage. In addition, fragility analyses using simplified models based on surveyed buildings reinforced these findings. The analyses suggested that RC frame and hybrid systems were insufficient in ensuring life safety during the earthquakes. Conversely, properly designed RC wall buildings are expected to perform well. This alignment between field observations and fragility analyses underscores the reliability of the study’s findings and emphasizes the effectiveness of RC wall construction in mitigating seismic risks and protecting life and property.

Decoding Continuous Tracking Eye Movements from Cortical Spiking Activity.

Eye movements are the primary way primates interact with the world. Understanding how the brain controls the eyes is therefore crucial for improving human health and designing visual rehabilitation devices. However, brain activity is challenging to decipher. Here, we leveraged machine learning algorithms to reconstruct tracking eye movements from high-resolution neuronal recordings. We found that continuous eye position could be decoded with high accuracy using spiking data from only a few dozen cortical neurons. We tested eight decoders and found that neural network models yielded the highest decoding accuracy. Simpler models performed well above chance with a substantial reduction in training time. We measured the impact of data quantity (e.g. number of neurons) and data format (e.g. bin width) on training time, inference time, and generalizability. Training models with more input data improved performance, as expected, but the format of the behavioral output was critical for emphasizing or omitting specific oculomotor events. Our results provide the first demonstration, to our knowledge, of continuously decoded eye movements across a large field of view. Our comprehensive investigation of predictive power and computational efficiency for common decoder architectures provides a much-needed foundation for future work on real-time gaze-tracking devices.

Simplified Gravity Load Collapse Dynamic Analysis of Old-Type Reinforced Concrete Frames

The results of shaking table tests from previous studies on a one-story, two-bay reinforced concrete frame—exhibiting both shear and axial failures—were compared with nonlinear dynamic analyses using simplified models intended to evaluate the collapse potential of older reinforced concrete structures. To replicate the nonlinear behavior of columns, whether shear-critical or primarily flexure-dominant, a one-component beam model was applied. This model features a linear elastic element connected in series to a rigid plastic, linearly hardening spring at each end, representing a concentrated plasticity component. To account for strength degradation through path-dependent plasticity, a negative slope model as degradation was implemented, linking points at both shear and axial failure. The shear failure points were determined through pushover analysis of shear-critical columns using Phaethon software. Although the simplified model provided a reasonable approximation of the overall frame response and lateral strength degradation, especially in terms of drift, its reduced computational demands led to some discrepancies between the calculated and measured shear forces and drifts during certain segments of the time-history response.

Ranking of power grid facilities as a tool for choosing a repair management system

In the long-term planning of repairs at power grid facilities in conditions of limited operational and investment resources, it is important to determine the "points of maximum return", that is, the objects whose repair will cause the greatest effect in terms of the company's target priorities. This requires an operational assessment and ranking of electric grid facilities, taking into account the technological and resource characteristics of the energy company. The article presents a decision-making methodology that, based on operational calculations and priority assessment, enables to identify power grid facilities with the highest repair efficiency in the sense of criteria characterizing the strategic goals of asset management of an energy company. The methodology combines methods of modeling and calculating the probabilities of failure in consumer power supply systems (PSS) with the determination of indices of the technical condition of equipment, methods for calculating labor costs for maintaining and restoring reliability and assessing damage from its disruptions.The developed methodology allows the energy company — the owner of assets, when choosing the most effective repair management system according to enlarged indicators and simplified models, to form prioritization lists of objects for which the critical repair rank corresponds to its strategic priorities. The possibility of cost estimation of the total labor costs for the maintenance of facilities for each of the received prioritization lists in comparison with the repair fund stated by the energy company expands the scope of the methodology and increases the validity of the decisions made. The methodology has been verified and confirmed by consideration on the example of a real PSS of oil fields using up-to-date operational information.

Carbon Nanotubes in Cement—A New Approach for Building Composites and Its Influence on Environmental Effect of Material

An addition of carbon nanostructures to cement paste is problematic due to the difficulties in obtaining homogenous mixtures. The paper reports on a more effective way of mixing carboxylated multi-walled carbon nanotubes (MWCNT-COOH) in cement pastes. The additional biological impact of the studied nanomodified cement was analyzed in the case of two moss species’ vitality. The applied approach of obtaining a homogeneous mixture is based on intense mechanochemical mixing of MWCNT-COOH together with polycarboxylate superplasticizer (SP). As a result, a more homogenous suspension of MWCNT-COOH within a liquid superplasticizer, suitable for addition to hydrophilic cement paste, was obtained. FT-IR/Raman spectroscopy was used for materials’ characterization. To explain the mixing process at the molecular level, systematic theoretical studies using density functional theory (DFT) were performed. The structures, interaction energies and IR/Raman vibrational spectra of model carboxylic acids, mixed with functionalized SWCNTs as simplified models of real MWCNTs, were obtained. Due to the controversial opinions on the environmental hazards of carbon nanostructures, additional in vivo studies were performed. In this case, effects of cement modified by the addition of small amounts of MWCNT-COOH with SP in comparison to the composite without carbon nanostructures and control subsoil on the vitality of mosses Polytrichum formosum and Pseudoscleropodium purum were studied.

Statistical-based modeling strategy for entrance skin dose estimation in patient undergoing body interventional radiology.

Interventional radiology (IR) provides significant advancements in diagnostic and therapeutic procedures, yet concerns persist regarding radiological risks such as erythema, burns, and epilation. Direct dose measurements observed difficulties regarding the perturbation of the detector probe in X-ray images during fluoroscopy-guided procedures, high-cost expenses, and non-compliant patients. This study aims to develop a statistical-based model for estimating entrance skin dose (ESD) in body IR procedures using patient radiation-dose recording data. Models are categorized into vascular and non-vascular procedures. This study demonstrates that the simplified models are sufficient in estimating patient ESDs for both IR groups, with a 95% confidence interval. This user-friendly method enables radiologists to calculate doses without complex parameters such as the backscatter factor and mass-energy absorption coefficient, as required in conventional calculation methods. It not only does this support to radiologists in effectively refining treatment protocols, but also enables patients to monitor their received doses immediately after treatment ends.

Full 3D thermal-hydraulic and electric modelling of quench propagation in HTS conductors

Abstract A fully three-dimensional multi-physics model to simulate quench propagation in high temperature superconducting (HTS) conductors for fusion applications is presented. It accounts for thermal, electric and fluid dynamics throughout the entire transient. The need for high-fidelity models for quench simulations comes from the bulky layouts of many HTS conductors that are being proposed. The model is then validated against experimental data, showing a good agreement on all the relevant quantities (local voltages and temperatures). It is shown that the detailed model improves the quality of the agreement with the measured data with respect to more simplified models. It also allows an insight on the temperature distributions in the conductor cross-section, which can be relevant for the interpretation of experimental data as well as to support the design of quench detection strategies which rely on local temperature variations.

Selective Glycan Presentation in Liquid‐Ordered or ‐Disordered Membrane Phases and its Effect on Lectin Binding

AbstractGlycan‐protein interactions play a key role in various biological processes from fertilization to infections. Many of these interactions take place at the glycocalyx—a heavily glycosylated layer at the cell surface. Despite its significance, studying the glycocalyx remains challenging due to its complex, dynamic, and heterogeneous nature. This study introduces a glycocalyx model allowing for the first time to control spatial organization and heterogeneity of the glycan moieties. Glycan‐mimetics with lipid‐moieties that partition into either liquid‐ordered (Lo, lipid rafts) or liquid‐disordered (Ld) phases of giant unilamellar vesicles (GUVs), which serve as simplified cell membrane models mimicking lipid rafts, are developed. This phase‐specific allocation allows controlled placement of glycan motifs in distinct membrane environments, creating heteromultivalent systems that replicate the natural glycocalyx's complexity. We show that phase localization of glycan mimetics significantly influences recruitment of protein receptors to the membrane. Glycan‐conjugates in the ordered phase demonstrate enhanced lectin binding, supporting the idea that raft‐like domains facilitate stronger receptor interactions. This study provides a platform for systematically investigating spatial and dynamic presentation of glycans in biological systems and presents the first experimental evidence that glycan accumulation in lipid rafts enhances receptor binding affinity, offering deeper insights into the glycocalyx‘s functional mechanisms.

Selective Glycan Presentation in Liquid-Ordered or -Disordered Membrane Phases and its Effect on Lectin Binding.

Glycan-protein interactions play a key role in various biological processes from fertilization to infections. Many of these interactions take place at the glycocalyx-a heavily glycosylated layer at the cell surface. Despite its significance, studying the glycocalyx remains challenging due to its complex, dynamic, and heterogeneous nature. This study introduces a glycocalyx model allowing for the first time to control spatial organization and heterogeneity of the glycan moieties. Glycan-mimetics with lipid-moieties that partition into either liquid-ordered (Lo, lipid rafts) or liquid-disordered (Ld) phases of giant unilamellar vesicles (GUVs), which serve as simplified cell membrane models mimicking lipid rafts, are developed. This phase-specific allocation allows controlled placement of glycan motifs in distinct membrane environments, creating heteromultivalent systems that replicate the natural glycocalyx's complexity. We show that phase localization of glycan mimetics significantly influences recruitment of protein receptors to the membrane. Glycan-conjugates in the ordered phase demonstrate enhanced lectin binding, supporting the idea that raft-like domains facilitate stronger receptor interactions. This study provides a platform for systematically investigating spatial and dynamic presentation of glycans in biological systems and presents the first experimental evidence that glycan accumulation in lipid rafts enhances receptor binding affinity, offering deeper insights into the glycocalyx's functional mechanisms.

Synergistic Modeling of Liquid Properties: Integrating Neural Network-Derived Molecular Features with Modified Kernel Models.

A significant challenge in applying machine learning to computational chemistry, particularly considering the growing complexity of contemporary machine learning models, is the scarcity of available experimental data. To address this issue, we introduce an approach that derives molecular features from an intricate neural network-based model and applies them to a simpler conventional machine learning model that is robust to overfitting. This method can be applied to predict various properties of a liquid system, including viscosity or surface tension, based on molecular features drawn from the ab initio calculated free energy of solvation. Furthermore, we propose a modified kernel model that includes Arrhenius temperature dependence to incorporate theoretical principles and diminish extreme nonlinearity in the model. The modified kernel model demonstrated significant improvements in certain scenarios and possible extensions to various theoretical concepts of molecular systems.

Dynamic plastic damage analysis of bending-torsion coupling in cross stiffener systems under mass impact

ABSTRACT This study proposes a simpler theoretical model to predict the bending-torsion coupling dynamic plastic response of cross stiffener systems under mass impact. Traditional methods often rely on torque balance and momentum conservation equations, which involve complex and difficult calculations. To simplify this process, the study uses the law of energy conservation and applies the variational method to solve the dynamic plastic damage of the cross stiffener system. The plastic response of a single stiffener is first analyzed and compared with existing theories, showing consistent results for maximum displacement. The study then extends the model to more complex conditions, performing comparative analyses on stiffeners of different sizes through numerical simulation. The findings show that the maximum displacement at the load and node remains within a reasonable range, with a displacement difference of only 0.9 mm and an error margin of 4.79% in one of the test conditions.

Assessment of Tube–Fin Contact Materials in Heat Exchangers: Guidelines for Simulation and Experiments

This paper describes experiments on finned tube heat exchangers, focusing on reducing the thermal contact resistance at the contact between the pipe and the lamella. Various contact materials, such as solders and adhesives, were investigated. Several methods of establishing contact were tested, including blowtorch soldering, brazing, and furnace soldering. Thermal camera measurements were carried out to assess the performance of the contact materials. Moreover, finite element analysis was performed to evaluate the contact materials and establish guidelines in the fin–tube connection modeling by comparing simplified models with the realistic model. Blowtorch brazing tests were successful while soldering attempts failed. During the thermographic measurements, reflective surfaces could be measured after applying a thin layer of paint with high emissivity. These measurements did not provide valuable results; thus, the contact materials were assessed using a finite element analysis. The results from the finite element analysis showed that all the inspected contact materials provided better heat transfer than not using a contact material. The heat transfer rate of the tight-fit realistic model was found to be 33.65 for air and 34.9 for the Zn-22Al contact material. This finding could be utilized in developing heat exchangers with higher heat transfer with the same size.

The Limits of Our Explanation: A Case Study in Myxococcus xanthus Cooperation

AbstractIn this article, I demonstrate two ways in which our major theories of the evolution of cooperation may fail to capture particular social phenomena. The first shortcoming of our current major theories stems from the possibility of mischaracterizing the cooperative problem in game theory. The second shortcoming of our current major theories is the insensitivity of these explanatory models to ecological and genomic context. As a case study to illustrate these points, I will use the cooperative interaction of a species of myxobacteria called Myxococcus xanthus. M. xanthus cooperate in many areas of their life cycle—in quorum sensing, social motility, fruiting body formation, and predation. I focus in particular on predation as we have not yet discovered an adequate explanation of how they sustain cooperative predation in the face of developmental cheats. In explaining why we have not, I draw generalizable conclusions that shed light on our use of simplified models to explain real-world behaviors in a variety of organisms.

Leveraging Massive MIMO for Enhanced Energy Efficiency in High-Density IoT Networks

Maximizing energy efficiency (EE) in massive multiple-input multiple-output (MIMO) systems, while supporting the rapid expansion of Internet of Things (IoT) devices, is a critical challenge. In this paper, we delve into the intricate operations geared toward enhancing EE in such complex environments. To effectively support a multitude of IoT devices, we adopt a strategy of heavy reference signal (RS) reuse, and in this circumstance, we formulate the EE metrics and their corresponding inverses to determine pivotal operational parameters. These EE-centric parameters encompass factors such as the number of service antennas in the base station (BS), the number of IoT devices, and permissible coverage extents. Our objective is to calibrate these parameters to meet a predefined EE threshold, ensuring optimal system performance. Additionally, we recognize the indispensable role of Peak-to-Average Power Ratio (PAPR) reduction techniques, particularly in multicarrier systems, to further enhance EE. As such, we employ clipping-based PAPR reduction methods to mitigate signal distortions and bolster overall efficiency. Theoretical EE metrics are derived based on formulated signal-to-interference-plus-noise ratios (SINRs), yielding insightful closed-form expressions for the operational parameters. Leveraging two distinct EE metric models, we undertake parameter determinations, accounting for the levels of approximation. Intriguingly, our analysis reveals that even simplified models exhibit remarkable applicability in real-world scenarios, with a minimal margin of error. The results not only underscore the practical applicability of our theoretical constructs but also highlight the potential for significant EE enhancements in massive MIMO systems, thereby contributing to sustainable evolution in the IoT era.

Performance Analysis of ML and DL Models: Impact of Linear and Non-Linear Optimizers on Model Efficiency

Introduction: Emphasizing the effect of linear and non-linear optimizers on the performance of machine learning (ML) and deep learning (DL) models, this work addresses the CIFAR-10 image classification task. The work explores over multiple data sets how optimizers influence model behavior and efficiency. Objectives: With a view on five primary assessment metrics—accuracy, precision, recall, F1-score, and AUC the main purpose is to evaluate on SVM and CNN models the performance of linear optimizers (gradient descent, SGD) and non-linear optimizers (Adam, RMSProp). Methods: The study offers fair and complete comparison by way of a consistent evaluation strategy. Analyzing how different optimizer influences model convergence rates, computational cost, and training stability takes front stage in the experimental setting. Designed on CIFAR-10, optimizers are used on ML and DL models whose performance on all metrics is noted. Results: Results show that non-linear optimizers especially Adam much increase CNN model performance by means of faster convergence, higher classification accuracy, and better model stability. While helpful for simpler models, linear optimizers such SGD show slower convergence and limited adaptation with more complicated data. Conclusions: This study provides direction on selecting appropriate optimizers depending on model complexity, job needs, and computing restrictions coupled with important information on optimizing model training in image classification and other areas.