Multilayer Perceptron Neural Network Research Articles

Measuring volume fractions of a three-phase flow without separation utilizing an approach based on artificial intelligence and capacitive sensors.

Many different kind of fluids in a wide variety of industries exist, such as two-phase and three-phase. Various combinations of them can be expected and gas-oil-water is one of the most common flows. Measuring the volume fraction of phases without separation is vital in many aspects, one of which is financial issues. Many methods are utilized to ascertain the volumetric proportion of each phase. Sensors based on measuring capacity are so popular because this kind of sensor operates seamlessly and autonomously without necessitating any form of segregation or disruption for measuring in the process. Besides, at the present moment, Artificial intelligence (AI) can be nominated as the most useful tool in several fields, and metering is no exception. Also, three main type of regimes can be found which are annular, stratified, and homogeneous. In this paper, volume fractions in a gas-oil-water three-phase homogeneous regime are measured. To accomplish this objective, an Artificial Neural Network (ANN) and a capacitance-based sensor are utilized. To train the presented network, an optimized sensor was implemented in the COMSOL Multiphysics software and after doing a lot of simulations, 231 different data are produced. Among all obtained results, 70 percent of them (161 data) are awarded to the train data, and the rest of them (70 data) are considered for the test data. This investigation proposes a new intelligent metering system based on the Multilayer Perceptron network (MLP) that can estimate a three-phase water-oil-gas fluid's water volume fraction precisely with a very low error. The obtained Mean Absolute Error (MAE) is equal to 1.66. This dedicates the presented predicting method's considerable accuracy. Moreover, this study was confined to homogeneous regime and cannot measure void fractions of other fluid types and this can be considered for future works. Besides, temperature and pressure changes which highly temper relative permittivity and density of the liquid inside the pipe can be considered for another future idea.

Performance assessment of packed bed systems for humidity control in greenhouse applications: An experimental-based AI modeling approach

Optimal humidity control is essential for enhancing crop yields and ensuring favorable growth conditions in greenhouse agriculture. Packed bed systems have emerged as effective tools for regulating humidity levels, yet accurately assessing their performance remains unexplored, especially for temperate oceanic climates. This paper presents a packed-bed system with water as the working fluid to increase the humidity level during the winter for greenhouse cultivation. Accordingly, an experimental setup is developed, and a detailed parametric study is conducted. Further, an artificial intelligence (AI)-based modeling approach is developed for evaluating the performance of packed bed systems for greenhouse applications under varying environmental conditions with various inlet air flowrates, such as 176 m³/hr, 286 m³/hr, 383 m³/hr, and 428 m³/hr. Operating with water at an average temperature of 15.7 °C and a flow rate of 12.8 kg/min, the system achieves a significant 50% increase in humidity ratio, transitioning from an inlet humidity ratio of 6 g/kgda to an outlet ratio of 9 g/kgda, indicating its efficiency in elevating air humidity levels. The multi-layer perceptron neural network, trained with 112 non-repeated datasets and employing a 2-10-10-1 topology, demonstrates high accuracy and resilience in estimating Δωa values for the packed bed system, with predictions closely aligning with experimental data and exhibiting a maximum discrepancy within ± 2.5%. This research contributes to the advancement of precision agriculture practices by providing a comprehensive framework for assessing and improving humidity control in greenhouse environments.

On the additive artificial intelligence-based discovery of nanoparticle neurodegenerative disease drug delivery systems.

Neurodegenerative diseases are characterized by slowly progressing neuronal cell death. Conventional drug treatment strategies often fail because of poor solubility, low bioavailability, and the inability of the drugs to effectively cross the blood-brain barrier. Therefore, the development of new neurodegenerative disease drugs (NDDs) requires immediate attention. Nanoparticle (NP) systems are of increasing interest for transporting NDDs to the central nervous system. However, discovering effective nanoparticle neuronal disease drug delivery systems (N2D3Ss) is challenging because of the vast number of combinations of NP and NDD compounds, as well as the various assays involved. Artificial intelligence/machine learning (AI/ML) algorithms have the potential to accelerate this process by predicting the most promising NDD and NP candidates for assaying. Nevertheless, the relatively limited amount of reported data on N2D3S activity compared to assayed NDDs makes AI/ML analysis challenging. In this work, the IFPTML technique, which combines information fusion (IF), perturbation theory (PT), and machine learning (ML), was employed to address this challenge. Initially, we conducted the fusion into a unified dataset comprising 4403 NDD assays from ChEMBL and 260 NP cytotoxicity assays from journal articles. Through a resampling process, three new working datasets were generated, each containing 500,000 cases. We utilized linear discriminant analysis (LDA) along with artificial neural network (ANN) algorithms, such as multilayer perceptron (MLP) and deep learning networks (DLN), to construct linear and non-linear IFPTML models. The IFPTML-LDA models exhibited sensitivity (Sn) and specificity (Sp) values in the range of 70% to 73% (>375,000 training cases) and 70% to 80% (>125,000 validation cases), respectively. In contrast, the IFPTML-MLP and IFPTML-DLN achieved Sn and Sp values in the range of 85% to 86% for both training and validation series. Additionally, IFPTML-ANN models showed an area under the receiver operating curve (AUROC) of approximately 0.93 to 0.95. These results indicate that the IFPTML models could serve as valuable tools in the design of drug delivery systems for neurosciences.

Machine-learning classifier models for predicting sarcopenia in the elderly based on physical factors.

As the size of the elderly population gradually increases, musculoskeletal disorders, such as sarcopenia, are increasing. Diagnostic techniques such as X-rays, computed tomography, and magnetic resonance imaging are used to predict and diagnose sarcopenia, and methods using machine learning are gradually increasing. This study aimed to create a model that can predict sarcopenia using physical characteristics and activity-related variables without medical diagnostic equipment, such as imaging equipment, for the elderly aged 60 years or older. A sarcopenia prediction model was constructed using public data obtained from the Korea National Health and Nutrition Examination Survey. Models were built using Logistic Regression, Support Vector Machine (SVM), XGBoost, LightGBM, RandomForest, and Multi-layer Perceptron Neural Network (MLP) algorithms, and the feature importance of the models trained with the algorithms, except for SVM and MLP, was analyzed. The sarcopenia prediction model built with the LightGBM algorithm achieved the highest test accuracy, of 0.848. In constructing the LightGBM model, physical characteristic variables such as body mass index, weight, and waist circumference showed high importance, and activity-related variables were also used in constructing the model. The sarcopenia prediction model, which consisted of only physical characteristics and activity-related factors, showed excellent performance. This model has the potential to assist in the early detection of sarcopenia in the elderly, especially in communities with limited access to medical resources or facilities. Geriatr Gerontol Int 2024; 24: 595-602.

A comparative study of neural network architectures for software vulnerability forecasting

Abstract The frequency of cyberattacks has been rapidly increasing in recent times, which is a significant concern. These attacks exploit vulnerabilities present in the software components that constitute the targeted system. Consequently, the number of vulnerabilities within these software components serves as an indicator of the system’s level of security and trustworthiness. This paper compares the accuracy, trainability and stability to configuration parameters of several neural network architectures, namely Long Short-Term Memory, Multilayer Perceptron and Convolutional Neural Network. These architectures are utilized for forecasting the number of software vulnerabilities within a specified timeframe for a specific software product. By evaluating these neural network models, our aim is to provide insights into their performance and effectiveness in vulnerability forecasting.

NCLDR: Nearest-Neighbor Contrastive Learning with Dual Correlation Loss for Dimensionality Reduction

Dimensionality reduction is an efficient method for alleviating the issue of dimensionality in high-dimensional data. As a popular self-supervised learning method, contrastive learning has recently garnered considerable attention. In this paper, we propose NCLDR: Nearest-Neighbor Contrastive Learning with Dual Correlation Loss for Dimensionality Reduction, a novel dimensionality reduction method that is porting a contrastive learning framework to the specific task of dimensionality reduction. Firstly, NCLDR uses the nearest-neighbor to construct feature pairs from the training set itself. Afterwards, to decorrelate feature variables that produce representations invariant across such pairs, a basic multi-layer perceptron (MLP) network architecture with a dual correlation loss function is designed. Compared to most dimensionality reduction methods, NCLDR bypasses the complexity of optimizing kNN graphs and facilitates the embedding of out-of-sample data. Additionally, it also alleviates the issue of “dimensional collapse” in the low-dimensional representation space. Finally, experimental results demonstrate that the proposed method achieves significant improvements over state-of-the-art dimensionality reduction methods.

Analysing LULC transformations using remote sensing data: insights from a multilayer perceptron neural network approach

ABSTRACT The study examines the complex dynamics of changes in LULC over three decades, focused on the years 1992, 2002, 2012, and 2022. The research highlights the significance of comprehending these alterations within the framework of environmental and socio-economic consequences. The changes in land use and land cover (LULC) have significant and far-reaching effects on ecosystems, biodiversity, and human livelihoods. This study offers useful information for politicians, conservationists, and urban planners by examining historical patterns and forecasting future changes. The study utilized a Multilayer Perceptron Neural Network (MLP-NN), a well-known machine learning technique that excels at collecting intricate patterns. This model’s design had three layers: input, hidden, and output. The model underwent 10,000 iterations during its training process, and a thorough statistical analysis was conducted to assess the impact of each driving component. The MLP-NN model demonstrated impressive performance, with a skill measure of 0.8724 and an accuracy rate of 89.08%. The accuracy of the LULC estimates for 2022 was verified by comparing them with observed data, ensuring the model’s reliability. Moreover, the presence of evidence likely was found to be a significant factor that had a substantial impact on the accuracy of the model. The study highlights the effectiveness of the MLP-NN model in accurately predicting changes in LULC. The model’s exceptional accuracy and proficiency make it a powerful tool for future LULC forecasts. Identifying the primary causes of model performance and understanding their implications may help to enhance land management strategies, encourage spatial planning, guide accurate decision-making, and facilitate the development of policies that align with sustainable growth and development.

Data-driven modeling of the aerodynamic deformation and drag for a freely moving drop in the sub-critical Weber number regime

Accurate prediction of the dynamics and deformation of freely moving drops is crucial for numerous droplet applications. When the Weber number is finite but below a critical value, the drop deviates from its spherical shape and deforms as it is accelerated by the gas stream. Since aerodynamic drag on the drop depends on its shape oscillation, accurately modeling the drop shape evolution is essential for predicting the drop’s velocity and position. In this study, 2D axisymmetric interface-resolved simulations were performed to provide a comprehensive dataset for developing a data-driven model. Parametric simulations were conducted by systematically varying the drop diameter and free-stream velocity, achieving wide ranges of Weber and Reynolds numbers. The instantaneous drop shapes obtained in simulations are characterized by spherical harmonics. Temporal data of the drag and modal coefficients are collected from the simulation data to train a Nonlinear Auto-Regressive models with eXogenous inputs (NARX) neural network model. The overall model consists of two multi-layer perceptron networks, which predict the modal coefficients and the drop drag, respectively. The drop shape can be reconstructed with the predicted modal coefficients. The model predictions are validated against the simulation data in the testing set, showing excellent agreement for the evolutions of both the drop shape and drag.

Advanced fusion of MTM-LSTM and MLP models for time series forecasting: An application for forecasting the solar radiation

Accurate time series forecasting has become increasingly important across various domains such as finance, energy, and medicine. This study introduces an innovative hybrid model that leverages the power of neural networks, precisely Many To Many LSTM (MTM LSTM) and Multilayer Perceptron (MLP), to improve time series forecasting accuracy. In this new combination, we trained network MTM LSTM to approximate the target at each step, and finally, we used network MLP to combine these approximations. To perform the evaluation, we made a forecast for the amount of solar energy radiation in the city of Mashhad, Iran. The experiment results concerning MSE and MAE showed that the proposed method with five lags outperforms the standard models.We hypothesize that MTM LSTM can effectively capture solar radiation's intricate temporal dependencies and nonlinearity. At the same time, MLP can enhance function approximation by modeling complex interactions, resulting in improved forecast accuracy. By employing the hybrid MTM LSTM and MLP model, we achieved improved accuracy in predicting solar energy radiation, which has significant implications for the renewable energy sector and its energy management and planning applications. This research advances time series forecasting techniques, highlighting the effectiveness of combining neural networks to address complex and dynamic patterns in time-dependent data.Overall, our findings underscore the potential and efficacy of the proposed hybrid model as a robust tool for accurate time series forecasting in various domains, supporting effective decision-making and planning processes.

A time-frequency-based interval decomposition ensemble method for forecasting gasoil prices under the trend of low-carbon development

Given that gasoil plays a crucial role in carbon emission reduction, in this paper we propose a time-frequency-based interval decomposition ensemble (TFIDE) learning approach to forecast gasoil prices and capture the nonlinear impact of the global trend of low-carbon development on gasoil prices. The proposed method integrates bivariate empirical mode decomposition (BEMD), an interval multilayer perceptron (IMLP) network and a threshold autoregressive interval (TARI) model. First, we use BEMD to decompose interval-valued weekly gasoil prices into a finite number of complex-valued intrinsic mode function (IMF) components and a residual component. Second, we apply the IMLP model to forecast the IMFs and the TARI model to predict the residual part with predictors of carbon reduction technology and carbon emission concerns. After that, we combine all the forecasting results to generate the final gasoil price interval forecasting results. Our empirical results show that our carbon reduction technology variable improves middle-frequency IMF forecasting and that carbon emission concerns have a nonlinear impact on long-term gasoil price intervals. Furthermore, the proposed TFIDE approach outperforms other competing methods under different accuracy measurements.

Dynamic viscosity of low GWP refrigerants in the liquid phase: An empirical equation and an artificial neural network

This study presents a simple correlation for describing the temperature and pressure dependence of the liquid dynamic viscosity of low GWP refrigerants, namely HydroFluoroOlefins (HFOs) and HydroChloroFluoroOlefins (HCFOs). The model has 3 input parameters (i.e., reduced temperature, reduced pressure, and acentric factor) and 6 coefficients which were regressed on 794 experimental data collated from the literature for 7 alternative refrigerants (i.e., R1233zd(E), R1234yf, R1234ze(E), R1234ze(Z), R1224yd(Z), R1336mzz(E), and R1336mzz(Z)). Moreover, a multi-layer perceptron neural network for the liquid dynamic viscosity of the studied fluids was developed from the selected dataset. The artificial network has the same 3 input parameters of the correlation and one hidden layer with 19 neurons. The results of the proposed correlation proved that it is an accurate model for calculating the dynamic viscosity of the studied liquid refrigerants, despite its simplicity. It ensured an average absolute relative deviation of the liquid dynamic viscosity (AARD(η)) of 2.88 %, lower than that given by other literature correlations. As expected, the multi-layer perceptron neural network provided the best results for all the selected refrigerants (AARD(η) = 0.86 % for the complete dataset), proving that it can be considered a reference for the development of other models.

A Comparative Analysis of Neural Network Architectures for Predicting Indian Rice Production

Rice (Oryza sativa) is one of the most important cereal crops in World and feeds more than a third of the world’s population. In Asian region, rice is a main source of nutrition and provides 30% to 70% of the daily calories for half of the world’s population. Here, in this study two different neural network models were used in prediction of rice production of India. It was observed that the accuracy score of Multi-layer perceptron neural network is better than Radial basis function in prediction of rice production. The loss/error value for Multi-layer perceptron (MLP) model is lower than Radial basis function (RBF) model. The relative error is found to be high for MLP.

Multilayer Perceptron Neural Network with Arithmetic Optimization Algorithm-Based Feature Selection for Cardiovascular Disease Prediction

In the healthcare field, diagnosing disease is the most concerning issue. Various diseases including cardiovascular diseases (CVDs) significantly influence illness or death. On the other hand, early and precise diagnosis of CVDs can decrease chances of death, resulting in a better and healthier life for patients. Researchers have used traditional machine learning (ML) techniques for CVD prediction and classification. However, many of them are inaccurate and time-consuming due to the unavailability of quality data including imbalanced samples, inefficient data preprocessing, and the existing selection criteria. These factors lead to an overfitting or bias issue towards a certain class label in the prediction model. Therefore, an intelligent system is needed which can accurately diagnose CVDs. We proposed an automated ML model for various kinds of CVD prediction and classification. Our prediction model consists of multiple steps. Firstly, a benchmark dataset is preprocessed using filter techniques. Secondly, a novel arithmetic optimization algorithm is implemented as a feature selection technique to select the best subset of features that influence the accuracy of the prediction model. Thirdly, a classification task is implemented using a multilayer perceptron neural network to classify the instances of the dataset into two class labels, determining whether they have a CVD or not. The proposed ML model is trained on the preprocessed data and then tested and validated. Furthermore, for the comparative analysis of the model, various performance evaluation metrics are calculated including overall accuracy, precision, recall, and F1-score. As a result, it has been observed that the proposed prediction model can achieve 88.89% accuracy, which is the highest in a comparison with the traditional ML techniques.

Can machine learning efficiently predict symmetry breaking in physical problems like Bose-Einstein condensates?

In this work, our objective is to evaluate whether machine learning algorithms combined with computational methods used in physical problems such as spontaneous symmetry breaking in Bose-Einstein condensates are capable of efficiently predicting results obtained from solutions of nonlinear equations. Thus, we use K-means algorithms, support vector machines (SVM) and artificial neural network (ANN) multilayer perceptron to evaluate the efficiency of each algorithm in the process of predicting theoretical results in regions where computational calculations cannot be performed. either by numerical inaccuracy or computational cost. The results show that sophisticated machine learning algorithms can predict the behavior of represented physical systems with a satisfactory quality, where we especially highlight the correct interpretation of the physical problem in border regions and/or transitions.

Predicting the friction angle of clays using a multi-layer perceptron neural network enhanced by yeo-johnson transformation and coral reefs optimization

The accurate prediction of the friction angle of clays is crucial for assessing slope stability in engineering applications. This study addresses the importance of estimating the friction angle and presents the development of four soft computing models: YJ-FPA-MLPnet, YJ-CRO-MLPnet, YJ-ACOC-MLPnet, and YJ-CSA-MLPnet. First of all, the Yeo-Johnson (YJ) transformation technique was used to stabilize the variance of data and make it more suitable for parametric statistical models that assume normality and equal variances. This technique is expected to improve the accuracy of friction angle prediction models. The friction angle prediction models then utilized multi-layer perceptron neural networks (MLPnet) and metaheuristic optimization algorithms to further enhance performance, including flower pollination algorithm (FPA), coral reefs optimization (CRO), ant colony optimization continuous (ACOC), and cuckoo search algorithm (CSA). The prediction models without the YJ technique, i.e. FPA-MLPnet, CRO-MLPnet, ACOC-MLPnet, and CSA-MLPnet, were then compared to those with the YJ technique, i.e. YJ-FPA-MLPnet, YJ-CRO-MLPnet, YJ-ACOC-MLPnet, and YJ-CSA-MLPnet. Among these, the YJ-CRO-MLPnet model demonstrated superior reliability, achieving an accuracy of up to 83% in predicting the friction angle of clay in practical engineering scenarios. This improvement is significant, as it represents an increase from 1.3% to approximately 20% compared to the models that did not utilize the YJ transformation technique.

Predictive analysis of concrete slump using a stochastic search-consolidated neural network

Attaining a dependable measurement of concrete slump is crucial as it is a valuable indication of concrete workability. On the other hand, complexities associated with costly traditional approaches have driven engineers to use indirect efficient models such as metaheuristic-based machine learning for approximating the slump. While the literature shows promising application of some metaheuristic techniques for this purpose, the large variety of these algorithms calls for evaluating the most capable ones to keep the solution updated. Stochastic fractal search (SFS) is one of the most powerful optimization algorithms in the literature that has not received appropriate attention in analyzing concrete mechanical parameters. In the present research, a multi-layer perceptron neural network (NN-MLP), is enhanced using the SFS. The proposed SFS-NN-MLP model aims to predict the slump based on the amount of ingredients in the mixture, as well as the curing age of specimens. Accuracy assessment revealed that the proposed model can deal with the assigned task with excellent accuracy. It indicates that the SFS could properly tune the parameters required for training the NN-MLP, and consequently, the trained network could reliably calculate the slump of specimens that were not analyzed before. For comparative validation, the SFS was replaced with two similar optimizers, namely elephant herding optimization algorithm (EHO) and slime mould algorithm (SMA). Based on the calculated mean square errors of 5.6526, 6.1129, and 7.3561 along with mean absolute errors of 4.6657, 5.0078, and 6.3066, as well as the percentage-Pearson correlation coefficients of 78.06 %, 73.95 %, and 58.11 %, respectively for the SFS-NN-MLP, EHO–NN–MLP, and SMA–NN–MLP, it was shown that the SFS-NN-MLP is the most accurate predictor. Hereupon, the SFS-NN-MLP model is recommended to be effectively used for obtaining a cost-efficient approximation of concrete slump in real-world projects.

Impact of Goods and Service Tax on MSME Sector: A Study Using Artificial Neural Network and Multivariate Analysis of Variance

Micro, small and medium enterprises (MSME) is one the vital sector of the developing economy. Changes in fiscal policy directly impact this sector the most. The present article aims to examine the impact of the recent Indian indirect system tax overhaul, that is, Goods and Service Tax (GST), on the MSME sector. We have employed multivariate analysis of variance and multi-layer perceptron neural network on the extensive data set for 60 months from April 2015 to March 2020. The empirical findings state that a significant increase has been observed in the MSMEs’ export, revenue generation and gross domestic product in the post-GST era. The findings may aid the emerging economies that are planning to alter tax systems to boost the MSME sector at the international level. JEL Codes: H20, L00, B22, C01, C45

Nonlinear hydrological time series modeling to forecast river level dynamics in the Rio Negro Uruguay basin.

Floods significantly impact the well-being and development of communities. Hence, understanding their causes and establishing methodologies for risk prevention is a critical challenge for effective warning systems. Complex systems such as hydrological basins are modeled through hydrological models that have been utilized to understand water recharge of aquifers, available volume of dams, and floods in diverse regions. Acquiring real-time hydrometeorological data from basins and rivers is vital for establishing data-driven-based models as tools for the prediction of river-level dynamics and for understanding its nonlinear behavior. This paper introduces a hydrological model based on a multilayer perceptron neural network as a useful tool for time series modeling and forecasting river levels in three stations of the Rio Negro basin in Uruguay. Daily time series of river levels and rainfall serve as the input data for the model. The assessment of the models is based on metrics such as the Nash-Sutcliffe coefficient, the root mean square error, percent bias, and volumetric efficiency. The outputs exhibit varying model performance and accuracy during the prediction period across different sub-basin scales, revealing the neural network's ability to learn river dynamics. Lagged time series analysis demonstrates the potential for chaos in river-level time series over extended time periods, mainly when predicting dam-related scenarios, which shows physical connections between the dynamical system and the data-based model such as the evolution of the system over time.

Integrated neural network and evolutionary algorithm approach for liver fibrosis staging: Can artificial intelligence reduce patient costs?

Staging liver fibrosis is important, and liver biopsy is the gold standard diagnostic tool. We aim to design and evaluate an artificial neural network (ANN) method by taking advantage of the Teaching Learning-Based Optimization (TLBO) algorithm for the prediction of liver fibrosis stage in blood donors and hepatitis C patients. We propose a method based on a selection of machine learning classification methods including multilayer perceptron (MLP) neural network, Naive Bayesian (NB), decision tree, and deep learning. Initially, the synthetic minority oversampling technique (SMOTE) is performed to address the imbalance in the dataset. Afterward, the integration of MLP and TLBO is implemented. We propose a novel algorithm that reduces the number of required patient features to seven inputs. The accuracy of MLP using 12 features is 0.903, while that of the proposed MLP with TLBO is 0.891. Besides, the diagnostic accuracy of all methods, except the model designed with the Bayesian network, increases when the SMOTE balancer is applied. The decision tree-based deep learning methods show the highest levels of accuracy with 12 features. Interestingly, with the use of TLBO and seven features, MLP reached an accuracy rate of 0.891, which is quite satisfactory when compared with those of similar studies. The proposed model provides high diagnostic accuracy, while reducing the required number of properties from the samples. The results of our study show that the recruited algorithm of our study is more straightforward, with a smaller number of required properties and similar accuracy.

A real-time solution method for three-dimensional steady temperature field of transformer windings based on mechanism-embedded cascade network

To enhance the computation efficiency and accuracy of three-dimensional steady temperature field of transformer windings, we propose a new non-invasive Reduced Order Model (ROM) based on a mechanism-embedded cascade network. Initially, a snapshot matrix is formed from the Full Order Model (FOM) and then combined with Proper Orthogonal Decomposition (POD) to extract key modal features that characterize the temperature field. Subsequently, a cascade network architecture, integrating Multilayer Perceptron (MLP) and Radial Basis Function Neural Network (RBFNN), is devised to swiftly map working condition parameters to modal coefficients. Additionally, the cascade network is embedded with condition sensitivity and modal contribution mechanisms to further enhance prediction accuracy. Finally, by linearly weighting the modes with predicted modal coefficients, a rapid reconstruction of the steady temperature field in transformer windings is achieved. Validation against Fluent software simulations and experimental measurements demonstrate a close agreement, with computational errors of less than 4K and an impressive single solution time of only 0.0087 s, which is 48760 times faster compared to Fluent software.