Multilayer Perceptron Algorithm Research Articles

An Analysis of Traditional Methods and Deep Learning Methods in SSVEP-Based BCI: A Survey

The brain–computer interface (BCI) is a direct communication channel between humans and machines that relies on the central nervous system. Neuroelectric signals are collected by placing electrodes, and after feature sampling and classification, they are converted into control signals to control external mechanical devices. BCIs based on steady-state visual evoked potential (SSVEP) have the advantages of high classification accuracy, fast information conduction rate, and relatively strong anti-interference ability, so they have been widely noticed and discussed. From k-nearest neighbor (KNN), multilayer perceptron (MLP), and support vector machine (SVM) classification algorithms to the current deep learning classification algorithms based on neural networks, a wide variety of discussions and analyses have been conducted by numerous researchers. This article summarizes more than 60 SSVEP- and BCI-related articles published between 2015 and 2023, and provides an in-depth research and analysis of SSVEP-BCI. The survey in this article can save a lot of time for scholars in understanding the progress of SSVEP-BCI research and deep learning, and it is an important guide for designing and selecting SSVEP-BCI classification algorithms.

A QSAR-based application for the prediction of lethal blood concentration of new psychoactive substances

The rapid development and introduction of new psychoactive substances (NPS) into illegal markets present an enormous challenge for forensic toxicologists, as there is limited knowledge about their toxicity in humans. To strengthen forensic interpretation of NPS intoxication cases, we have developed a predictive model for estimating human lethal blood concentrations (LBC) of various NPS. This quantitative structure-activity relationship (QSAR) model focuses on opioids, designer benzodiazepines, synthetic cathinones, synthetic cannabinoids, and phenethylamines. Utilising linear regression and multilayer perceptron algorithms, the models was trained using data from the existing literature. A toxicological significance-based approach have been applied to refine the selection of training data. The model demonstrated satisfactory performance metrics through cross-validation (R ≈ 0.8, MAE ≈ 0.6) and comparison with experimental data (R ≈ 0.9). A Python-based web application have been developed to facilite the use of the created model in predicting LBC of NPS. Despite the model's reliability, limitations due to data availability, quality and the complexities of post-mortem toxicology mean that its predictions should be interpreted with caution.

Prediction of local site influence on seismic vulnerability using machine learning: A study of the 6 February 2023 Türkiye earthquakes

This study uses machine learning to analyze local seismic features' influence on damage from the 6 February 2023 Türkiye Earthquakes.The input features include Vs30 (the average shear wave velocity to a depth of 30 m), f0 (the predominant frequency of the site), A0 (HVSR ratio for the site), and EBd (engineering bedrock depth), along with the target feature of damage status for 44 locations. Machine learning involves Random Forest (RF), K-nearest Neighbor (KNN), Logistic Regression (LR), Decision Trees (DT), Support Vector Machines (SVM), Stochastic Gradient Descent (SGD), and Multilayer Perceptron (MP) algorithms. Also, five-fold cross-validation is employed to acquire suitable hyperparameters, enhancing its efficacy in modeling small sample sets. RF emerged as the most effective in whole performance metrics, presenting recall scores for damage and no damage conditions respectively by a 94% and 92% ratio and achieving a damage status prediction accuracy of 93%. All remaining algorithms also exhibited remarkable performance, reaching a minimum accuracy of 89% by DT, and recall score for no damage condition with 80% by MP and damage condition with 88% by SVM and SGD. The outcomes definitively designate EBd as the most crucial parameter, attributing 52% importance to its role in building damage occurrence within the study area. In contrast, significance values were determined as 24%, 18%, and 6% for f0, Vs30 and A0 respectively. These findings underscore the importance of demonstrating that initial damage estimation in high seismic hazard zones can be effectively carried out using machine learning approaches through seismic-based local site parameters.

Rapid detection and quantification of melamine, urea, sucrose, water, and milk powder adulteration in pasteurized milk using Fourier transform infrared (FTIR) spectroscopy coupled with modern statistical machine learning algorithms

There is an evident requirement for a rapid, efficient, and simple method to screen the authenticity of milk products in the market. Fourier transform infrared (FTIR) spectroscopy stands out as a promising solution. This work employed FTIR spectroscopy and modern statistical machine learning algorithms for the identification and quantification of pasteurized milk adulteration. Comparative results demonstrate modern statistical machine learning algorithms will improve the ability of FTIR spectroscopy to predict milk adulteration compared to partial least square (PLS). To discern the types of substances utilized in milk adulteration, a top-performing multiclassification model was established using multi-layer perceptron (MLP) algorithm, delivering an impressive prediction accuracy of 97.4 %. For quantification purposes, bayesian regularized neural networks (BRNN) provided the best results for the determination of both melamine, urea and milk powder adulteration, while extreme gradient boosting (XGB) and projection pursuit regression (PPR) gave better results in predicting sucrose and water adulteration levels, respectively. The regression models provided suitable predictive accuracy with the ratio of performance to deviation (RPD) values higher than 3. The proposed methodology proved to be a cost-effective and fast tool for screening the authenticity of pasteurized milk in the market.

Comparative models of support-vector machine, multilayer perceptron, and decision tree ‎predication approaches for landslide ‎susceptibility analysis

Abstract Naqadeh Region (NR) is one of the most sensitive regions regarding geo-hazards ‎occurrence in Northwest of Iran. The landslides triggering parameters that ‎identified for the studied region are classified as elevation, aspect, slope angle, ‎lithology, drainage density, distance to river, weathering, land-cover, ‎precipitation, vegetation, distance to faults, distance to roads, and distance to ‎the cities. These triggering factors are selected based on conducting field ‎survey, remote-sensing investigation, and historical development background ‎assessment. Regarding the investigations, 12 large-scale, 15 medium-scale, and 30 small-scale historical landslides ‎(57 in total) were recorded in the NR. The historical landslides were used to provide ‎sensitive area with high probability of ground movements. The objectives of this study are multifaceted, aiming to address critical gaps in understanding and predicting landslide susceptibility in the NR. First, the study seeks to evaluate and compare the effectiveness of ‎support-vector machine (SVM), multilayer perceptron (MLP), and decision tree ‎‎(DT) algorithms in predicting landslide susceptibility. So, as methodology, the ‎presented study used comparative models for landslide susceptibility based on ‎SVM, MLP, and DT approaches. The predictive models were compared based on model ‎accuracy as the area under the curve of the receiver operating characteristic ‎curve. According to the estimated results, MLP is the highest rank of overall ‎accuracy to provide susceptibility maps for landslides in NR. From a perspective of ‎the risk ability, the west and south-west sides of the county were identified within ‎the hazard area.

Blast furnace operation profile monitoring system based on DB-Mini-KMeans and multi-layer perceptron

This article proposes a system for monitoring a blast furnace operation based on DB-Mini-KMeans and multi-layer perceptron (MLP), which solves a difficult problem categorising the furnace operation types. The system uses the correlation between cooling wall temperature and furnace operation type, creates an furnace operation type recognition model with optimised K-Means and MLP algorithms and makes a comparison test with other recognition models; evaluates all kinds of operation types by combining six blast furnace production indexes, such as gas utilisation rate and airflow and establishes a monitoring system of furnace operation types in blast furnaces. The clustering performance of the system has been improved by nearly 40%, and the recognition accuracy of various operation types has reached more than 90%, covering functions such as two-dimensional schematic diagrams of furnace operation types and status warnings, which provide help and support for people to analyse furnace conditions.

Suicide attempts and non-suicidal self-injury in Chinese adolescents: Predictive models using a neural network model

IntroductionSuicide attempts (SA) are a significant contributor to suicide deaths, and non-suicidal self-injury (NSSI) can increase the risk of SA. Many adolescents experience both NSSI and SA, which are affected by various factors. This study aimed to identify the risk factors and essential warning signs of SA, establish a predictive model for SA using multiple dimensions and large samples, and provide a multidimensional perspective for clinical diagnosis and intervention. MethodsA total of 9140 participants aged 12–18 years participated in an online survey; 6959 participants were included in the statistical analysis. A multilayer perceptron algorithm was used to establish a prediction model for adolescent SA (with or without); adolescents with NSSI behavior were extracted as a subgroup to establish a prediction model. ResultsBoth the prediction model performance of the SA group and the NSSI-SA subgroup were strong, with high accuracy, and AUC values of 0.93 and 0.88, indicating good discrimination. Decision curve analysis (DCA) demonstrated that the clinical intervention value of the prediction results was high and that the clinical intervention benefits of the NSSI-SA subgroup were greater than those of the SA group. ConclusionsOur study demonstrated that the predictive model has a high degree of accuracy and discrimination, thereby identifying significant factors associated with adolescent SA. As long as adolescents exhibit NSSI behavior, relative suicide interventions should be implemented to prevent future hazards. This study can provide guidance and more nuanced insights for clinical diagnosis as well as a foundation for clinical treatment.

Concomitant Prediction of the Ki67 and PIT-1 Expression in Pituitary Adenoma Using Different Radiomics Models.

To preoperatively predict the high expression of Ki67 and positive pituitary transcription factor 1 (PIT-1) simultaneously in pituitary adenoma (PA) using three different radiomics models. A total of 247 patients with PA (training set: n = 198; test set: n = 49) were included in this retrospective study. The imaging features were extracted from preoperative contrast-enhanced T1WI (T1CE), T1-weighted imaging (T1WI), and T2-weighted imaging (T2WI). Feature selection was performed using Spearman's rank correlation coefficient and least absolute shrinkage and selection operator (LASSO). The classic machine learning (CML), deep learning (DL), and deep learning radiomics (DLR) models were constructed using logistic regression (LR), support vector machine (SVM), and multi-layer perceptron (MLP) algorithms. The area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, specificity, accuracy, negative predictive value (NPV) and positive predictive value (PPV) were calculated for the training and test sets. In addition, combined with clinical characteristics, the best CML and the best DL models (SVM classifier), the DL radiomics nomogram (DLRN) was constructed to aid clinical decision-making. Seven CML features, 96 DL features, and 107 DLR features were selected to construct CML, DL and DLR models. Compared to CML and DL model, the DLR model had the best performance. The AUC, sensitivity, specificity, accuracy, NPV and PPV were 0.827, 0.792, 0.800, 0.796, 0.800 and 0.792 in the test set, respectively. Compared with CML and DL models, the DLR model shows the best performance in predicting the Ki67 and PIT-1 expression in PAs simultaneously.

Indoor air quality prediction modeling for a naturally ventilated fitness building using RNN-LSTM artificial neural networks

Indoor air quality prediction modeling for a naturally ventilated fitness building using RNN-LSTM artificial neural networks

MLP and RBF Algorithms in Finance: Predicting and Classifying Stock Prices amidst Economic Policy Uncertainty

In the realm of stock market prediction and classification, the use of machine learning algorithms has gained significant attention. In this study, we explore the application of Multilayer Perceptron (MLP) and Radial Basis Function (RBF) algorithms in predicting and classifying stock prices, specifically amidst economic policy uncertainty. Stock market fluctuations are greatly influenced by economic policies implemented by governments and central banks. These policies can create uncertainty and volatility, which in turn makes accurate predictions and classifications of stock prices more challenging. By leveraging MLP and RBF algorithms, we aim to develop models that can effectively navigate these uncertainties and provide valuable insights to investors and financial analysts. The MLP algorithm, based on artificial neural networks, is able to learn complex patterns and relationships within financial data. The RBF algorithm, on the other hand, utilizes radial basis functions to capture non-linear relationships and identify hidden patterns within the data. By combining these algorithms, we aim to enhance the accuracy of stock price prediction and classification models. The results showed that both MLB and RBF predicted stock prices well for a group of countries using an index reflecting the impact of news on economic policy and expectations, where the MLB algorithm proved its ability to predict chain data. Countries were also classified according to stock price data and uncertainty in economic policy, allowing us to determine the best country to invest in according to the data. The uncertainty surrounding economic policy is what makes stock price forecasting so crucial. Investors must consider the degree of economic policy uncertainty and how it affects asset prices when deciding how to allocate their assets.

A hybrid multilayerperceptron-extremegradientboost approach for precise state of charge and state of health assessment

Lithium ion batteries are popularly used in various energy storage systems. Accurate battery's State of Charge (SoC) and State of Health (SoH) estimate is essential for energy storage system's safety. This manuscript proposes a hybrid approach for estimating SoC and SoH of lithium ion battery. The proposed hybrid technique is the joint execution of both the Multi-Layer Perceptron (MLP) with Extreme Gradient boosting (XGBoost) algorithm. Hence, it is named as MLP-XG Boost Algorithm. The major objective of the proposed system is to improve the safety of energy storage systems and to achieve the highly accurate SoH. The MLP algorithm is utilized to analyze and trained the battery data. The XG Boosting algorithm is utilized to preprocessed and estimate the SoC and SoH of a lithium-ion battery. The proposed method's effectiveness is assessed using the MATLAB platform and contrasted with other methods that are currently in use. In every system now in use, including Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN), the proposed approach produces better results. From the result it is concluded that the proposed method based Mean Square Error and Mean Absolute Error Root is less than 1 at various temperatures such as 0∘C, 25∘Cand 45∘C and it is better and efficient.

Acoustic impedance prediction based on extended seismic attributes using multilayer perceptron, random forest, and extra tree regressor algorithms

Acoustic impedance is the product of the density of a material and the speed at which an acoustic wave travels through it. Understanding this relationship is essential because low acoustic impedance values are closely associated with high porosity, facilitating the accumulation of more hydrocarbons. In this study, we estimate the acoustic impedance based on nine different inputs of seismic attributes in addition to depth and two-way travel time using three supervised machine learning models, namely extra tree regression (ETR), random forest regression, and a multilayer perceptron regression algorithm using the scikit-learn library. Our results show that the R2 of multilayer perceptron regression is 0.85, which is close to what has been reported in recent studies. However, the ETR method outperformed those reported in the literature in terms of the mean absolute error, mean squared error, and root-mean-squared error. The novelty of this study lies in achieving more accurate predictions of acoustic impedance for exploration.

Empowering an Acute Kidney Injury 3D Graphene-Based Sensor Using Extreme Learning Machine.

This study reports on the application of an extreme learning machine (ELM) in near-real-time kidney monitoring via urine neutrophil gelatinase-associated lipocalin (NGAL) detection with a 3D graphene electrode. This integration marks the first instance of combining a graphene-based electrode with machine learning to enhance the NGAL detection accuracy, building on our group's 2020 research. The methodology involves two key components: a graphene electrode functionalized with a lipocalin-2 antibody for NGAL detection and the ELM application for improved prediction accuracy by using urine analysis data. The results show a significant 15% increase in the area under the curve (AUC) for NGAL determination, with error reduction from ±6 to 0.54 ng/mL within a linear range of 2.7-140 ng/mL. The ELM also lowered the detection limit from 14.8 to 0.89 ng/mL and increased accuracy, precision, sensitivity, specificity, and F1 score for AKI prediction by 8.89, 30.69, 6.78, 9.94, and 19.07%, respectively. These findings underscore the efficacy of simple neural networks in enhancing graphene-based electrochemical sensors for AKI biomarkers. ELM was chosen for its optimal performance-resource balance, with a comparative analysis of ELM, support vector machines, multilayer perceptron, and random forest algorithms also included. This research suggests the potential for miniaturizing AI-enhanced sensors for practical applications.

Towards a Framework for Performance Management and Machine Learning in a Higher Education Institution

Towards a Framework for Performance Management and Machine Learning in a Higher Education Institution

Monitoring of Plant Ecological Units Cover Dynamics in a Semiarid Landscape from Past to Future Using Multi-Layer Perceptron and Markov Chain Model

Anthropogenic activities and natural disturbances cause changes in natural ecosystems, leading to altered Plant Ecological Units (PEUs). Despite a long history of land use and land cover change detection, the creation of change detection maps of PEUs remains problematic, especially in arid and semiarid landscape. This study aimed to determine and describe the changes in PEUs patterns in the past and present, and also predict and monitor future PEUs dynamics using the multi-layer perceptron-Markov chain (MLP-MC) model in a semiarid landscape in Central Zagros, Iran. Analysis of PEUs classification maps formed the basis for the identification of the main drivers in PEUs changes. First, an optimal time-series dataset of Landsat images were selected to derive PEUs classification maps in three periods, each separated by 16 years. Then, PEUs multi-temporal maps classified for period 1 (years 1986–1988) period 2 (years 2002–2004), and period 3 (years 2018–2020) were employed to analyze and predict PEUs dynamics. The dominant transitions were identified, and the transition potential was determined by developing twelve sub-models in the final change prediction process. Transitions were modeled using a Multi-Layer Perceptron (MLP) algorithm. To predict the PEU map for period 3, two PEUs classification maps of period 1 and period 2 were used using the MLP-MC method. The classified map and the predicted map of period 3 were used to evaluate and validate the predicted results. Finally, based on the results, transitions of future PEUs were predicted for the year 2036. The MLP-MC model proved to be a powerful model that can predict future PEUs dynamics that are the result of current human and managerial activities. The findings of this study demonstrate that the impact of anthropogenic processes and management activities will become visible in the natural environment and ecosystem in less than a decade.

Performance analysis of a blockchain enabled domestic power consumption monitoring and forecasting system

Purpose The purpose of this paper is to develop a blockchain-based data capture and transmission system that will collect real-time power consumption data from a household electrical appliance and transfer it securely to a local server for energy analytics such as forecasting. Design/methodology/approach The data capture system is composed of two current transformer (CT) sensors connected to two different electrical appliances. The CT sensors send the power readings to two Arduino microcontrollers which in turn connect to a Raspberry-Pi for aggregating the data. Blockchain is then enabled onto the Raspberry-Pi through a Java API so that the data are transmitted securely to a server. The server provides real-time visualization of the data as well as prediction using the multi-layer perceptron (MLP) and long short term memory (LSTM) algorithms. Findings The results for the blockchain analysis demonstrate that when the data readings are transmitted in smaller blocks, the security is much greater as compared with blocks of larger size. To assess the accuracy of the prediction algorithms data were collected for a 20 min interval to train the model and the algorithms were evaluated using the sliding window approach. The mean average percentage error (MAPE) was used to assess the accuracy of the algorithms and a MAPE of 1.62% and 1.99% was obtained for the LSTM and MLP algorithms, respectively. Originality/value A detailed performance analysis of the blockchain-based transmission model using time complexity, throughput and latency as well as energy forecasting has been performed.

Utilizing Deep Neural Networks for Chrysanthemum Leaf and Flower Feature Recognition

Chrysanthemums, a significant genus within the Asteraceae, hold a paramount position in the global floricultural industry, second only to roses in market demand. The proliferation of diverse chrysanthemum cultivars presents a formidable challenge for accurate identification, exacerbated by the abundance of varieties, intricate floral structures, diverse floret types, and complex genetic profiles. Precise recognition of chrysanthemum phenotypes is indispensable to navigating these complexities. Traditional methods, including morphology studies, statistical analyses, and molecular markers, have fallen short due to their manual nature and time-intensive processes. This study presents an innovative solution employing deep learning techniques for image-based chrysanthemum phenotype recognition. Leveraging machine learning, our system autonomously extracts key features from chrysanthemum images, converting morphological data into accessible two-dimensional representations. We utilized Support Vector Machine (SVM) and Multilayer Perceptron (MLP) algorithms to construct frameworks for processing image data and classifying chrysanthemum cultivars based on color, shape, and texture. Experimental results, encompassing 10 cultivars, 10 flower colors, and five flower shapes, consistently demonstrated recognition accuracy ranging from 79.29% up to 97.86%. This tool promises streamlined identification of flower traits, and we anticipate its potential for real-time identification enhancements in future iterations, promising advances in chrysanthemum cultivation and exportation processes. Our approach offers a novel and efficient means to address the challenges posed by the vast diversity within chrysanthemum species, facilitating improved management, breeding, and marketing strategies in the floricultural industry.

Machine learning-based virtual screening of multi-target anti-obesity compounds from medicinal and edible plants: A combined in silico and in vitro study

In response to the limited effectiveness of existing weight loss food products, we sought to apply machine learning-based virtual screening methods to identify potential anti-obesity functional compounds from medicinal and edible plants and validate their in vitro activities. Firstly, we construct and evaluate the machine learning (ML) screening models using Multilayer Perceptron (MLP) and Random Forest (RF) algorithms. The receiver operating characteristic (ROC) curve demonstrates the high accuracy of MLP and RF models in screening for obese-related targets PL (pancreatic lipase) and AMPK (Adenosine 5′-monophosphate activated protein kinase). Subsequently, the tested ML models are employed to screen the constructed database, and Gypenoside LXVI (GYP) and alisol-b-23-acetate (ALI) are identified as compounds exhibiting favorable activity against both targets. The hit compounds are tested for their impact on lipase activity and lipid accumulation. The test results show that GYP and ALI have favorable inhibitory effects on pancreatic lipase (PL), with IC50 of 359.7 and 433.8 μg/mL. Furthermore, both GYP and ALI significantly reduced cellular lipid accumulation by 72.89% and 79.01% with the concentration increase to 40 μg/mL. The molecular docking results indicate that GYP and ALI can interact with several amino acid residues on the two target proteins, thereby affecting the activity of the target proteins. In conclusion, GYP and ALI can prevent and alleviate obesity by inhibiting PL activity and regulating AMPK signaling factors. We innovatively applied virtual screening based on ML to discover functional factors in food for anti-obesity purposes. This novel computational screening technique holds significant potential in the development of dietary supplements to combat obesity.

Utilization of machine learning models in predicting caries risk groups and oral health-related risk factors in adults

BackgroundThe aim of this study was to analyse the risk factors that affect oral health in adults and to evaluate the success of different machine learning algorithms in predicting these risk factors.MethodsThis study included 2000 patients aged 18 years and older who were admitted to the Department of Oral and Maxillofacial Radiology, Faculty of Dentistry, Gaziantep University, between September and December 2023. In this study, patients completed a 30-item questionnaire designed to assess the factors that affect the decayed, missing, and filled teeth (DMFT). Clinical and radiological examinations were performed, and DMFT scores were calculated after completion of the questionnaire. The obtained data were randomly divided into a 75% training group and a 25% test group. The preprocessed dataset was analysed using various machine learning algorithms, including naive Bayes, logistic regression, support vector machine, decision tree, random forest and Multilayer Perceptron algorithms. Pearson's correlation test was also conducted to assess the correlation between participants' DMFT scores and oral health risk factors. The performance of each algorithm was evaluated to determine the most appropriate algorithm, and model performance was assessed using accuracy, precision, recall and F1 score on the test dataset.ResultsA statistically significant difference was found between various factors and DMFT-based risk groups (p < 0.05), including age, sex, body mass index, tooth brushing frequency, socioeconomic status, employment status, education level, marital status, hypertension, diabetes status, renal disease status, consumption of sugary snacks, dry mouth status and screen time. When considering machine learning algorithms for risk group assessments, the Multilayer Perceptron model demonstrated the highest level of success, achieving an accuracy of 95.8%, an F1-score of 96%, and precision and recall rates of 96%.ConclusionsCaries risk assessment using a simple questionnaire can identify individuals at risk of dental caries, determine the key risk factors, provide information to help reduce the risk of dental caries over time and ensure follow-up. In addition, it is extremely important to apply effective preventive treatments and to prevent the general health problems that are caused by the deterioration of oral health. The results of this study show the potential of machine learning algorithms for predicting caries risk groups, and these algorithms are promising for future studies.

Dynamic Fractional-Order Grey Prediction Model with GWO and MLP for Forecasting Overseas Talent Mobility in China

International students play a crucial role in China’s talent development strategy. Thus, predicting overseas talent mobility is essential for formulating scientifically reasonable talent introduction policies, optimizing talent cultivation systems, and fostering international talent cooperation. In this study, we proposed a novel fractional-order grey model based on the Multi-Layer Perceptron (MLP) and Grey Wolf Optimizer (GWO) algorithm to forecast the movement of overseas talent, namely MGDFGM(1,1). Compared to the traditional grey model FGM(1,1), which utilizes the same fractional order at all time points, the proposed MGDFGM(1,1) model dynamically adjusts the fractional-order values based on the time point. This dynamic adjustment enables our model to better capture the changing trends in the data, thereby enhancing the model’s fitting capability. To validate the effectiveness of the MGDFGM(1,1) model, we primarily utilize Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) as the evaluation criteria for the prediction accuracy, as well as standard deviation (STD) as an indicator of the model stability. Furthermore, we perform experimental analysis to evaluate the predictive performance of the MGDFGM(1,1) model in comparison to NAÏVE, ARIMA, GM(1,1), FGM(1,1), LSSVR, MLP, and LSTM. The research findings demonstrate that the MGDFGM(1,1) model achieves a remarkably high level of prediction accuracy and stability for forecasting overseas talent mobility in China. The implications of this study offer valuable insights and assistance to government departments involved in overseas talent management.