Comparative Modeling Study Research Articles

Financial fraud detection: A comparative study of quantum machine learning models

In this research, a comparative study of four Quantum Machine Learning (QML) models was conducted for fraud detection in finance. We proved that the Quantum Support Vector Classifier model achieved the highest performance, with F1 scores of [Formula: see text] for fraud and nonfraud classes. Other models like the Variational Quantum Classifier (VQC), Estimator Quantum Neural Network (QNN), and Sampler QNN demonstrate promising results, propelling the potential of QML classification for financial applications. While they exhibit certain limitations, the insights attained pave the way for future enhancements and optimization strategies. However, challenges exist, including the need for more efficient quantum algorithms and larger and more complex datasets. This paper provides solutions to overcome current limitations and contributes new insights to the field of QML in fraud detection, with important implications for its future development.

Humeral head size predicts baseplate lateralization in reverse shoulder arthroplasty: a comparative computer model study

BackgroundIn reverse shoulder arthroplasty (RSA), the ideal combination of baseplate lateralization (BL), glenosphere size (GS) and overhang (GOH) with a commonly used 145° neck shaft angle (NSA) is unclear. This is the first study evaluating correlations of body height (BH), humeral head size (HS), glenoid height (GH) and association of gender with best glenoid configurations for range of motion (ROM) maintaining anatomic lateralization (aLAT) for optimized muscle length in 145° and less distalized 135° RSA. MethodsIn this computer model study, 22 CTs without joint narrowing were analyzed (11 male/female). A standardized semi-inlay 145° platform stem was combined with 20 glenoid configurations (baseplate (B) 25, 25+3/+6 lateralized (l), 29, 29+3/6l combined with glenosphere36, 36+2eccentric (e), 36+3l, 39, 39+3e, 39+3l , 42, 42+4e). Abduction-adduction, flexion-extension, external rotation-internal rotation, total ROM (TROM) and total notching relevant (TNR) ROM were computed, best TROM models respecting aLAT (-1mm to +1mm) and HS/GH recorded. Secondly, the 145° models (Ascend Flex stem; Stryker, Kalamazoo, MI, USA) were converted and compared to a 135° inlay RSA (New Perform stem; Stryker, Kalamazoo, MI, USA) maintaining GOH (6.5-7mm) and aLAT. ResultsBest 145° models had eccentric glenospheres (mean BL: 3.5mm, GOH 8.8mm, GS 38.1mm, distalization (DIS) 23mm). The 135° models had concentric glenospheres, mean BL 3.8mm, GOH 6.9mm, GS 39.7mm, DIS 14.1mm. Head size showed the strongest positive correlation with BL in 145° and 135° models (0.65/0.79). Despite reduced GOH in smaller females with a 135° NSA, adduction, external rotation, extension, TNR ROM and TROM were significantly increased (p=0.02, p=0.005, p=0.005, p=0.004, p=0.003), abduction however reduced (p=0.02). The same trends were seen for males. ConclusionsHead size is a practical measure in surgery or preoperatively, and the strong positive correlation with BL is a useful planning aid. Despite reduction of GOH, conversion to a less distalized 135° NSA is powerful to maintain and even significantly increase all components of TNR ROM (extension/external rotation/adduction) in small females with the drawback of reduced abduction which may however be compensated by scapula motion. Lateralization with a less distalized 135° RSA optimizes muscle length, may facilitate subscapularis repair and maintains highest rigid body motion.

A Comparative Study of Fractional-Order Models for Supercapacitors in Electric Vehicles

Fractional-order models display many advantages compared with integer-order equivalent circuit models for modeling energy storage system, such as more precision, fewer parameters. This paper carries out a comparative study of fractional-order equivalent circuit models for modeling supercapacitor. These models are fractional-order RC model, fractional-order classic model, fractional-order RRCW model, fractional-order dynamic model, fractional-order transmission line model. An experiment load platform is built to test the supercapacitor for collecting the feature data. To be fair, the parameters of all models are optimized to achieve their best fitting accuracy by using the genetic algorithm. Accuracy and complexity two indices are considered to evaluate the above five models based on different test cycles. Simulation results show that the fractional-order RRCW model has the ideally comprehensive performance.

A Comparative Study of Machine Learning Models for Early Detection of Skin Cancer Using Convolutional Neural Networks

A Comparative Study of Machine Learning Models for Early Detection of Skin Cancer Using Convolutional Neural Networks

Comparative Analysis of Turbulence Models for Thermal-Hydraulic Simulations in Aqueous Homogeneous Reactors

This article presents a comparative study of various turbulence models applied in the context of thermal-hydraulic simulations for liquid fuel reactors, specifically Aqueous Homogeneous Reactors (AHR) using Computational Fluid Dynamics. The objective was to assess the suitability of the turbulence models by comparing their results with data obtained from Large Eddy Simulation (LES). For that purpose, was compared the flow behavior predicted using the k-ε, SST, GEKO, DES, SBES, and LES turbulence models. The calculations were carried out in a simplified computational model derived from a pre-existing three-dimensional AHR conceptual design. By utilizing this simplified model, the study aimed to focus on the computational differences between the turbulence models, while minimizing the influence of other factors. The calculation results revealed that the k-ε model exhibited significant discrepancies with the LES, with relative differences for the fuel solution maximum temperature reaching up to 75 %. Among the remaining RANS models, the Shear Stress Transport (SST) model demonstrated the best compromise between accuracy and computational efficiency, with differences below 5 % and requiring only 1/5th of the time, compared to the LES model. The Scale-Resolving Simulation (SRS) models,DES and SBES, provided a more comprehensive description of flow behavior and results closer to LES, albeit with higher computational demands. Between these two models, only the DES model exhibited relative differences below or equal to 1 % compared to the LES model for the studied thermohydraulic parameters.

Comparative study of neural network and model averaging methods in nuclear β-decay half-life predictions

Nuclear β-decay half-lives are investigated using the two-hidden-layer neural network and compared with the model averaging method. By carefully designing the input and hidden layers of the neural network, the neural network achieves better accuracy of nuclear β-decay half-life predictions and well eliminates the too strong odd–even staggering predicted by the previous neural networks. For nuclei with half-lives less than 1 s, the neural network can describe experimental half-lives within 1.6 times. The half-life predictions of the neural network are further tested with the newly measured half-lives, demonstrating its reliable extrapolation ability not far from the training region. Compared to the model averaging method, the neural network has higher accuracy and smaller uncertainties of half-life predictions in the known region. When extrapolated to the unknown region, the half-life uncertainties of the neural network are still smaller than those of the model averaging method within about 5–10 steps for nuclei with 35 ≲ Z ≲ 90, while the model averaging method has smaller half-life uncertainties for nuclei near the drip line.

Improved predictions of permeability properties in cement-based materials: A comparative study of pore size distribution-based models

Permeability is a fundamental physical property of a cement-based material and is closely related to its pore structure. This study developed an improved model for permeability prediction of cementitious materials with a function of the pore size distribution (PSD), in which PSD is simulated from single and compound lognormal distribution functions, respectively. The efficiency of the improved model is demonstrated by calculating permeability for mortars and pastes available in the literature. It is found that, first, for cementitious materials with unimodal PSD, the single lognormal distribution function is adequate to describe the pore structures. In contrast, for cementing mixtures presenting multimodal PSD, the compounded lognormal distribution function is highly recommended to cover a wide range of pore components. Second, the permeability model based on the compound PSD can predict well permeability to gas intrusion, whereas the single PSD-based model is applicable for calculating permeability via water intrusion. Furthermore, slippage effect and water saturation degree are introduced in the PSD-based model. Results prove that the corrected gas permeability model from the compound PSD performs well in estimating the gas permeability in unsaturated states. Finally, the compound PSD-based model is extended by considering varying porosities and relative humidities, which can be used to quantitively describe permeability when cementitious materials undergoing pore structure degradation induced by durability issues.

A comparative study of wormhole models in f(R,T) gravity

We have conducted research on the presence of wormholes characterized by a logarithmic shape function within the framework of exponential [Formula: see text] gravity. This gravity model is defined as [Formula: see text], with [Formula: see text] representing an arbitrary constant. We find that the logarithmic shape function serves as expected, providing all the necessary conditions for traversable and asymptotically flat wormholes. We have examined the solutions with three distinct sets of physical constraints placed on [Formula: see text], [Formula: see text] and [Formula: see text]. viz. Model-I: [Formula: see text] with k being variable dependent on r and [Formula: see text] with [Formula: see text] being EoS parameter, Model-II: [Formula: see text] and [Formula: see text] with [Formula: see text] and [Formula: see text] being EoS parameters, Model-III: Trace is zero i.e. [Formula: see text] or [Formula: see text] i.e. [Formula: see text] gravity model. From Model-I, we have obtained a new parametrization for EoS parameter [Formula: see text] and from Model-III, we have obtained a new form of shape function [Formula: see text] in [Formula: see text] gravity theory.

Analysis of the severity of transport vehicle accidents by a comparative study of machine learning models

Analysis of the severity of transport vehicle accidents by a comparative study of machine learning models

A comparative study of YOLO models and a transformer-based YOLOv5 model for mass detection in mammograms

A comparative study of YOLO models and a transformer-based YOLOv5 model for mass detection in mammograms

Input-parameter optimization using a SVR based ensemble model to predict landslide displacements in a reservoir area – A comparative study

For reservoir landslides with limited in situ monitoring data, improving the accuracy of an ensemble prediction model through historical displacement-related factor deconstruction and model optimization is challenging. Unfortunately, even though various support vector regression (SVR) based ensemble prediction models have been proposed for this purpose, the optimal combination of new algorithms and their degree of improved prediction accuracy is still unclear. Based on four evaluation indicators, this paper presents a comparative study of a typical landslide displacement prediction model containing four important nodes (landslide displacement decomposition, inducing factor frequency-component extraction, input parameter selection, and optimization algorithm selection) vital to prediction accuracy. The empirical mode decomposition (EMD) series model was adopted to decompose the landslide displacement and extract the frequency component of the factors that affect the landslide movement. Four trajectory similarity judgment models were used to select input variables for the SVR-based prediction model. Nine swarm intelligence (SI) algorithms were used to help optimize the landslide prediction model. A case study shows that the SVR-based ensemble landslide prediction model works well in predicting the displacement of a slow-moving landslide. The combination of EEMD-CEEMDAN-LCSS-PSO yields the best prediction performance, with MAPE, RMSE, MAE, and R2 values of 0.004, 2.38, 8.64, and 0.96, respectively. The findings of this study can impact further studies of landslide ensemble prediction model optimization with insights into the algorithm selection, parameter setting, evaluation measures, and experimental settings.

A comparative study of combustion models for simulating partially premixed swirling natural gas flames

Combustion models are important for simulating chemical reaction properties and the turbulence-chemistry interaction (TCI) in combustion. However, little attention has been paid to the temperature/species distributions/velocity flow fields predicted by different combustion models and their requirements for mesh resolution. In the present work, a partially premixed swirling natural gas flame was separately simulated by two commonly-used combustion models in Reynolds-Averaged-Numerical-Simulation (RANS) simulations: the flamelet generated manifold (FGM) model and the eddy dissipation concept (EDC) model. It is found that the EDC model requires a higher quality of mesh than the FGM model from the temperature and species distributions perspective. It is possible for the EDC model to overpredict temperature, however, such a problem can be solved by improving the mesh quality. The combustion models do not differ obviously for the mesh refinement for flow fields. The further analysis find that the EDC model brings forward the reaction location (11 mm upstream than the FGM model). The weakened turbulence mixing and early chemical reactions make the EDC model derive discrepancies with measured values, resulting from the variations between its modelling principles and actual applications in the software ANSYS Fluent. The low turbulence viscosity makes the EDC model flame thicker. Overall, the FGM model showed better suitability and accuracy for such a partially-premixed swirling natural gas flame. This work connects the two common combustion models and partially premixed swirling flame characteristics in the RANS simulations, and highlights the effects on the temperature/species concentration/turbulence flow fields, in concerns of mesh resolution and reaction properties rather than sole applications, which can provide foundation for subsequent researchers.

A comparative study of machine learning regression models for predicting construction duration

ABSTRACT Over the past few decades, the construction industry has been suffering from project delays. It remains a recognized challenge to accurately predict the actual project’s progress, despite the vast amount of field data and scheduling methods available. In this paper, four different machine learning (ML) models including K-nearest neighbor (KNN), support vector regression (SVR), gradient boosting trees (GBT), and artificial neural network (ANN) were used for forecasting the construction duration of work areas. Field data were collected, including progress, workload, labor, weather, planned days, and location, as inputs and construction days as output for each work area to develop the models. A simple and widely used baseline model was used to confirm the accuracy and generalization capability of ML models. The comparison of the results revealed that ANN and GBT models produced significantly superior results. Additionally, the GBT model is substantially more computationally efficient. In conclusion, the GBT model could be successfully and effectively applied to improve the prediction of construction duration. The proposed ML models could be utilized as a decision support tool for construction project managers to adjust construction site resource allocation.

A Comparative Study of Conventional Transition Model and Reverse Transition Model for Wireless and Radar Applications

Microstrip line-to-waveguide transitions (MLWT) finds various applications in the field of RADAR engineering, navigation science, medical electronics, household equipment, satellite tracking equipment, wireless communications etc. Conventionally, the MLWT is designed so that the microstrip line is inserted in E-plane and perpendicular to the waveguide’s side wall, while the microstrip probe’s radiating face is towards the waveguide’s open end port. In this paper, an MLWT has been designed for Ka-band applications and the transition has been designed in such a way that the radiating face of the inserted microstrip line probe is towards the short circuit end of the waveguide. The method is called the reverse transition method. The microstrip signal reflected from the short circuit end comes in additive phase with the traveling wave and the strength of the traveling wave increases. MLWT has been developed using this method and results for Ka-band applications are presented in this paper. This work reported more than 20 dB improvement in return loss (RL), 0.1dB improvement in insertion loss (IL) and almost 7 to 8% enhancement in 10dB fractional bandwidth compared to the conventional method. Comparative analysis of the two methods has been shown in tabular form at the end of the paper.

Enhancing the selective electrical activation of human vagal nerve fibers: a comparative computational modeling study with validation in a rat sciatic model

Objective. Vagus nerve stimulation (VNS) is an emerging treatment option for a myriad of medical disorders, where the method of delivering electrical pulses can vary depending on the clinical indication. In this study, we investigated the relative effectiveness of electrically activating the cervical vagus nerve among three different approaches: nerve cuff electrode stimulation (NCES), transcutaneous electrical nerve stimulation (TENS), and enhanced TENS (eTENS). The objectives were to characterize factors that influenced nerve activation and to compare the nerve recruitment properties as a function of nerve fiber diameter. Methods. The Finite Element Model, based on data from the Visible Human Project, was implemented in COMSOL. The three simulation types were compared under a range of vertical and horizontal displacements relative to the location of the vagus nerve. Monopolar anodic stimulation was examined, along with latency and activation of different fiber sizes. Nerve activation was determined via the activating function and McIntyre-Richardson-Grill models, and activation thresholds were validated in an in-vivo rodent model. Results. While NCES produced the lowest activation thresholds, eTENS generally performed superior to TENS under the range of conditions and fiber diameters, producing activation thresholds up to three times lower than TENS. eTENS also preserved its enhancement when surface electrodes were displaced away from the nerve. Anodic stimulation revealed an inhibitory region that removed eTENS benefits. eTENS also outperformed TENS by up to four times when targeting smaller diameter nerve fibers, scaling similar to a cuff electrode. In latency and activation of smaller diameter nerve fibers, eTENS results resembled those of NCES more than a TENS electrode. Activation threshold ratios were consistent in in-vivo validation. Significance. Our findings expand upon previously identified mechanisms for eTENS and further demonstrate how eTENS emulates a nerve cuff electrode to achieve lower activation thresholds. This work further characterizes considerations required for VNS under the three stimulation methods.

A comparative study of the LiFePO4 battery voltage models under grid energy storage operation

Lithium iron phosphate (LFP) batteries are widely used in energy storage systems (EESs). In energy storage scenarios, establishing an accurate voltage model for LFP batteries is crucial for the management of EESs. This study has established three energy storage working conditions, including power fluctuation smoothing, peak shaving, and frequency regulation. Four voltage models for commercial LFP batteries are developed, including the second-order resistor-capacitor equivalent circuit model, hysteresis voltage reconstruction model (HVRM), one-state hysteresis voltage model, and back-propagation neural network model. To evaluate model suitability in energy storage working conditions, we compare terminal voltage simulation accuracy, SOC estimation accuracy using the extended Kalman filter algorithm, and SOC calculation time. Overall, among the four models, the HVRM proves more suitable for energy storage scenarios, offering guidance for selecting an LFP voltage model in such conditions. Using the hysteresis model, we analyze the hysteresis open-circuit voltage (OCV) variations of LFP batteries in three energy storage scenarios. Research findings indicate that under frequency regulation, OCV exhibits high-frequency, small-amplitude variations, while under power fluctuation smoothing and peak shaving scenarios, OCV takes on a cyclic pattern. Considering the hysteresis OCV enhances Kalman gain matching, thereby improving SOC estimation accuracy.

Enhancing lead adsorption capacity prediction in biochar: a comparative study of machine learning models and parameter optimization.

Machine learning models for predicting lead adsorption in biochar, based on preparation features, are currently lacking in the environmental field. Existing conventional models suffer from accuracy limitations. This study addresses these challenges by developing back-propagation neural network (BPNN) and random forest (RF) models using selected features: preparation temperature (T), specific surface area (BET), relative carbon content (C), molar ratios of hydrogen to carbon (H/C), oxygen to carbon (O/C), nitrogen to carbon (N/C), and cation exchange capacity (CEC). The RF model outperforms BPNN, improving R2 by 10%. Additional features and particle swarm optimization enhance the RF model's accuracy, resulting in an 8.3% improvement in R2, a decrease in RMSE by up to 56.1%, and a 55.7% reduction in MAE. The importance ranking of features places CEC > C > BET > O/C > H/C > N/C > T, highlighting the significance of CEC in lead adsorption. Strengthening the complexation effect may improve lead removal in biochar. This study contributes valuable insights for predicting and optimizing lead adsorption in biochar, addressing the accuracy gap in existing models. It lays the foundation for future investigations and the development of effective biochar-based solutions for sustainable lead removal in water remediation.

Enhancing daily rainfall prediction in urban areas: a comparative study of hybrid artificial intelligence models with optimization algorithms

Forecasting precipitation is a crucial input to hydrological models and hydrological event management. Accurate forecasts minimize the impact of extreme events on communities and infrastructure by providing timely and reliable information. In this study, six artificial intelligent hybrid models are developed to predict daily rainfall in urban areas by combining the firefly optimization algorithm (FA), invasive weed optimization algorithm (IWO), genetic particle swarm optimization algorithm (GAPSO), neural network (ANN), group method of data handling (GMDH), and wavelet transformation. Optimization algorithms increase forecasting accuracy by controlling all stages. A variety of criteria are used for validating the models, including correlation coefficient (R), root-mean-square error (RMSE), mean absolute error (MAE), critical success index (CSI), probability of detection (POD), and false alarm ratio (FAR). The proposed models are also evaluated in an urban area in Ahvaz, Iran. The GAPSO-Wavelet-ANN model is superior to other models for predicting daily rainfall, with an RMSE of 1.42 mm and an R of 0.9715.

Unraveling pine wilt disease: Comparative study of stochastic and deterministic model using spectral method

The pinewood nematode, which causes a disease known as pine wilt, is a serious threat to pine forests all over the world. In this paper, we analyze the dynamics of pine wilt disease and its effects on forest ecosystems using mathematical modeling of pine wilt disease, its numerical simulations, and analytical techniques. Understanding the nematode’s spread, evaluating the efficiency of management efforts, and foreseeing the long-term effects on pine populations are some of our research’s goals. To gain a comprehensive understanding of the dynamics of the disease, we categorize pine trees and vectors systematically into susceptible, exposed, and infected groups. We first presented disease dynamics through a flow diagram which is then transformed to system of ordinary differential equations along with non-negative initial conditions. The deterministic model is then transformed to a stochastic model. Using both deterministic and stochastic modeling approaches, this study explores the complex transmission dynamics of Pine Wilt disease. Basic reproduction number has been calculated via next generation technique and some cases of stability analysis on the basis of the values of reproduction number has been discussed. For the solution of stochastic model Legendre Spectral collocation method has been used as it is known for its high accuracy in approximating solutions to complex stochastic models. It can achieve exponential convergence rates, making it suitable for problems where high precision is required. The results of both deterministic and stochastic models are compared via different figures. We find crucial thresholds for disease transmission and highlight significant factors affecting the spread of diseases through simulations. Our research shows that preventing the severe impacts of pine wilt disease requires early detection and intervention. Additionally, our mathematical model is a useful tool for enhancing disease management plans. This research advances our knowledge of the dynamics of the pine wilt disease and provides useful advice for managing forests and promoting conservation. Our research highlights the need for quick action to protect pine forests from this harmful virus.

Dam Break Flow: A Comparative Model Study Using OpenFOAM and BASEMENT

Dam Break Flow: A Comparative Model Study Using OpenFOAM and BASEMENT