Kernel Function Research Articles

Development of a machine learning model related to explore the association between heavy metal exposure and alveolar bone loss among US adults utilizing SHAP: a study based on NHANES 2015-2018.

Alveolar bone loss (ABL) is common in modern society. Heavy metal exposure is usually considered to be a risk factor for ABL. Some studies revealed a positive trend found between urinary heavy metals and periodontitis using multiple logistic regression and Bayesian kernel machine regression. Overfitting using kernel function, long calculation period, the definition of prior distribution and lack of rank of heavy metal will affect the performance of the statistical model. Optimal model on this topic still remains controversy. This study aimed: (1) to develop an algorithm for exploring the association between heavy metal exposure and ABL; (2) filter the actual causal variables and investigate how heavy metals were associated with ABL; and (3) identify the potential risk factors for ABL. Data were collected from National Health and Nutrition Examination Survey (NHANES) between 2015 and 2018 to develop a machine learning (ML) model. Feature selection was performed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression with 10-fold cross-validation. The selected data were balanced using the Synthetic Minority Oversampling Technique (SMOTE) and divided into a training set and testing set at a 3:1 ratio. Logistic Regression (LR), Support Vector Machines (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), Decision Tree (DT), and XGboost were used to construct the ML model. Accuracy, Area Under the Receiver Operating Characteristic Curve (AUC), Precision, Recall, and F1 score were used to select the optimal model for further analysis. The contribution of the variables to the ML model was explained using the Shapley Additive Explanations (SHAP) method. RF showed the best performance in exploring the association between heavy metal exposure and ABL, with an AUC (0.88), accuracy (0.78), precision (0.76), recall (0.83), and F1 score (0.79). Age was the most important factor in the ML model (mean| SHAP value| = 0.09), and Cd was the primary contributor. Sex had little effect on the ML model contribution. In this study, RF showed superior performance compared with the other five algorithms. Among the 12 heavy metals, Cd was the most important factor in the ML model. The relationship of Co & Pb and ABL are weaker than that of Cd. Among all the independent variables, age was considered the most important factor for this model. As for PIR, low-income participants present association with ABL. Mexican American and Non-Hispanic White show low association with ABL compared to Non-Hispanic Black and other races. Gender feature demonstrates a weak association with ABL. In the future, more advanced algorithms should be developed to validate these results and related parameters can be tuned to improve the accuracy of the model. not applicable.

Investigation of the Gaussian Process with Various Kernel Functions for the Prediction of the Compressive Strength of Concrete

The Compressive Strength of Concrete (CSC) is a critical parameter for evaluating the quality of concrete used in various construction projects, including buildings, bridges, and roads. The primary objective of this study is to examine the efficacy of a Gaussian Process (GP) Machine Learning (ML) model employing two kernel functions: Radial Basis Function (RBF) and Polynomial (POL), for predicting the CSC, considering readily quantifiable parameters. Based on these kernel functions, two models were created for this prediction, GP-RBF and GP-POL. The modeling process employed a total of 369 concrete sample data, including compressive strength values and eleven other physico-mechanical properties, collected from the Cua Luc bridge project in Vietnam. This dataset was partitioned into a training set (70%) and a testing set (30%) for model training and validation. Various validation metrics, including R2, Root Mean Square Error (RMSE), and Mean Absolute Error (MAE), were used to evaluate and compare the models. The findings of this study demonstrated that both models GP-RBF and GP-POL exhibited strong performance in predicting CSC, with GP-POL demonstrating marginal superiority over GP-RBF. Consequently, it can be concluded that POL is more efficacious than RBF in training the GP model for CSC prediction.

A spatiotemporal autoregressive neural network interpolation method for discrete environmental factors

The spatiotemporal interpolation model is necessary for generating continuous distributions for spatiotemporally discrete sampling points. However, there remain challenges in spatiotemporal interpolation due to the complex spatiotemporal effect and the imprecise kernel functions. Here, we proposed a spatiotemporal autoregressive neural network interpolation model (STARNN) that incorporates adaptive spatiotemporal distance quantification and supervised learning. The 10-fold cross-validation modelling on sea surface temperature and coastal nutrients demonstrated that the STARNN model performs better than baseline models and can well depict reasonable spatiotemporal distributions for environmental factors. By proposing two stacked neural networks, the STARNN model can accurately integrate spatial and temporal distances and avoids subjective selection of the kernel function. This study developed a novel interpolation model for processing discrete spatiotemporal points by following the data-driven paradigm, which can offer decision support for simulating the spread of sea temperature anomalies and optimizing the distribution of water quality measurement stations.

Enhancing internet of things attack detection using principal component analysis and kernel principal component analysis with cosine distance and sigmoid kernel

The widespread adoption of internet of things (IoT) devices has brought about unprecedented levels of connectivity and convenience. However, it has also introduced significant challenges, particularly in the areas of security and privacy. This study addresses the critical issue of intrusion detection within IoT environments, with a specific focus on analyzing the Iot-23 dataset. Our methodology involves employing principal component analysis (PCA) and kernel PCA for dimensionality reduction. Subsequently, we utilize the k-nearest neighbors (KNN) algorithm for classification purposes. To optimize the performance of the KNN algorithm, we experiment with various feature scaling techniques, such as StandardScaler, MinMaxScaler, and RobustScaler, utilizing different distance metrics. In our analysis, we discovered that employing the cosine distance metric in combination with KNN resulted in superior intrusion detection performance when utilizing PCA. Additionally, when utilizing kernel PCA, we evaluated multiple kernel functions and determined that the radial basis function and sigmoid kernel yielded the most favorable results.

A nonlinear sensitivity matrix for improving the imaging quality in electrical resistance tomography

Electrical Resistance Tomography (ERT) can image the conductivity distribution in a detection field by boundary measurements. The ERT imaging refers to a typical nonlinear inverse problem, and currently it is represented by a linear sensitivity matrix. However, the existing linearized sensitivity matrix cannot effectively represent the nonlinear effect in the imaging process, and leads to low-quality ERT image. This paper combines an integral equation on ERT with the Geselowitz-Lehr formula, and applies a Gaussian kernel function to derive a nonlinear sensitivity matrix. The new matrix can effectively reflect both linear and nonlinear characteristics in ERT, and has a clear theoretical basis and rationality. Essentially, the limitation of the existing linear sensitivity matrix is greatly addressed. Compared to four typical ERT imaging algorithms with the linear sensitivity matrix, the use of nonlinear sensitivity matrix can averagely decrease the relative error to 0.16, while can increase the correlation coefficient to 0.88 along with various noise levels. Hence, the proposed nonlinear sensitivity matrix can effectively improve ERT imaging quality.

Optimizing Demand Forecasting Method with Support Vector Regression for Improved Inventory Planning

Problems arising from suboptimal production planning can cause inventory management to be less effective and efficient in the company. The lack of integrated presentation of information also causes less efficiency in making decisions. This study aims to obtain the best kernel function forecasting model by predicting ground rod sales using the Support Vector Regression (SVR) method in order to determine the level of forecasting accuracy and the results of ground rod forecasting in the future which are presented in an optimal data visualization. This problem-solving is done with the Support Vector Regression method, which consists of linear kernel functions, polynomial kernel functions, and radial basis function (RBF) kernel functions with the Grid Search Algorithm. Based on the results of the best parameter search that has been done using the grid search algorithm, it can be concluded that the best kernel function forecasting model is a linear kernel function with a value of C = 100 and ε = 10-3. The accuracy of this forecasting model has a MAPE value of training data and testing data of 2.048% and 1.569%, where this value is the smallest MAPE value compared to the MAPE value of the other two functions. After getting the best model, forecasting was carried out within five months, obtaining an average of 6,647 monthly pieces. The results of forecasting and historical sales are reviewed in a visualization of Business Intelligence data so that it is well exposed, where the forecasting shows an increase from every month.

Rational Approximation by Reproducing Kernel for Bergman Functions on Siegel Domains

ABSTRACTBased on recent developments of reproducing kernels in several complex variables, we study the kernel approximation for weighted Bergman functions on Siegel domains. To overcome the difficulties arising in Siegel domains, a new method is proposed in the proof of boundary vanishing property (BVP) at infinity by adopting the biholomorphic invariance property of weighted kernel functions. Once the BVP is verified, optimal parameters are selected and the corresponding orthogonal functions are constructed by induction according to the maximal selection principle.

Positive-Definite Converging Kernel Estimation of Long-Run Variance

Kernel estimators have been popular for decades in long-run variance estimation. To minimize the loss of efficiency measured by the mean-squared error in important aspects of kernel estimation, we propose a novel class of converging kernel estimators with three major properties: (a) the optimal bandwidth choice is model-free; (b) positive-definiteness is ensured through a principle-driven aggregation technique with no loss of theoretical efficiency; and (c) potentially misspecified prewhitening models and transformations of the time series do not harm the asymptotic efficiency. A shrinkage prewhitening transformation is proposed for more robust finite-sample performance. The estimator has a positive bias that diminishes with the sample size so that it is more conservative compared with the typically negatively biased classical estimators. The proposal improves upon standard kernel functions and can be well generalized to the multivariate case. We discuss its performance through simulation results and a real-data application in the forecast breakdown test.

A Reservoir Dam Monitoring Technology Integrating Improved ABC Algorithm and SVM Algorithm

A reservoir dam is a water conservancy project with large investment and high social and economic benefits, which plays an irreplaceable role in flood control, power generation, water storage, and urban water supply. There is a risk of accidents in the process of reservoir dams, so dam monitoring is an important means to achieve the safe operation of reservoirs. In this paper, taking advantage of the high-dimensional and nonlinear characteristics of dam monitoring data samples, the fusion-improved ABC (artificial bee colony) algorithm is introduced, and the SVM (support vector machine) algorithm is used to optimize the penalty factor and kernel function parameters. The test results of the ABC and SVM algorithm are relatively stable, with small fluctuation amplitude, which can continuously monitor water level, pore water pressure, dam deformation, temperature, humidity, vibration, and other indicators is less than 10%, which is significantly lower than the standard ABC algorithm, the standard ANN algorithm, and the standard SVM algorithm. The independence and characteristics of the ABC–SVM algorithm are significantly higher, and the correlation is 0.03, the RMS (root mean square) is 0.2334, which is lower than that of the standard ABC algorithm of 0.09, and the standard ANN algorithm of 0.8. The stability of the results and performance stability are analyzed, which is greater than 90%. The ABC and SVM is used to predict the displacement and deformation law of the reservoir dam.

Innovative segmentation technique for aerial power lines via amplitude stretching transform

Accurate segmentation of power line targets helps quickly locate faults, evaluate line conditions, and provides key image data support and analysis for the safe and stable operation of the power system.The aerial power line in segmentation due to the target is small, and the imaging reflected energy is weak, so the Unmanned Aerial Vehicle (UAV) aerial power line image is very susceptible to the interference of the environment line elements and noise, resulting in the detection of the power line target in the image of the defective, intermittent, straight line interferences and other low accuracy and real-time efficiency is not high. For this reason, this paper designs a pure amplitude stretching kernel function to form a Fourier amplitude vector field and uses this amplitude vector field to implement the stretching transformation of the amplitude field of the aerial power line image, so that the angular field after the Fourier inverse transformation can better react to the spatial domain line targets, and finally, after the Relative Total Variation (RTV) processing, the power line can be well detected. The proposed algorithm is compared with the main power line segmentation algorithms, such as Region Convolutional Neural Networks(R-CNN) and Phase Stretch Transform(PST). The average values of evaluation indicators PPA, MMPA and MMIoU of the image segmentation results of the proposed algorithm reach 0.96, 0.96 and 0.95 respectively, and the average time lag of detection is less than 0.2s, indicating that the accuracy and real-time performance of the segmentation results of the proposed algorithm are significantly better than those of the above algorithms.

Modified kernel global-local marginal fisher analysis for rolling bearing feature extraction

Abstract The vibration signals of rolling bearings usually exhibit high-dimensional, nonlinear, and non-Gaussian distribution characteristics due to long-term operation under complex working conditions. Therefore, we proposed a novel algorithm named Modified Kernel Global-Local Marginal Fisher Analysis (MKGLMFA) for bearing feature extraction and dimensionality reduction. The proposed MKGLMFA algorithm introduces the kernel function to map data into a high-dimensional space to represent data nonlinearly first. It enhances the within-class compactness and between-class dispersibility by considering spatial relationships and label information when constructing adjacency graphs and simultaneously exploits the local and global geometry of data. Furthermore, a bearing fault diagnosis approach is presented based on MKGLMFA. It first processes the original vibration signals through MKGLMFA to obtain low-dimensional manifold features. Then these characteristics were input into the K-nearest neighbor (KNN) classifier to achieve fault pattern recognition. The superiority of the proposed MKGLMFA algorithm in feature extraction is verified in comparison with some existing state-of-the-art machine learning methods on three rolling bearings datasets. And the subsequent classification diagnosis experiments indicate the effectiveness and high efficiency of the newly raised MKGLMFA algorithm. In comparison with the representative diagnosis methods, the proposed method can extract more sensitive discriminant features, and the classification accuracy of diagnosis is significantly improved in consequence.

Study on the Impact of LDA Preprocessing on Pig Face Identification with SVM.

In this study, the implementation of traditional machine learning models in the intelligent management of swine is explored, focusing on the impact of LDA preprocessing on pig facial recognition using an SVM. Through experimental analysis, the kernel functions for two testing protocols, one utilizing an SVM exclusively and the other employing a combination of LDA and an SVM, were identified as polynomial and RBF, both with coefficients of 0.03. Individual identification tests conducted on 10 pigs demonstrated that the enhanced protocol improved identification accuracy from 83.66% to 86.30%. Additionally, the training and testing durations were reduced to 0.7% and 0.3% of the original times, respectively. These findings suggest that LDA preprocessing significantly enhances the efficiency of individual pig identification using an SVM, providing empirical evidence for the deployment of SVM classifiers in mobile and embedded systems.

Fault Diagnosis and Optimized Operation and Maintenance of Drop-out Fuses in Alpine Regions Using MKL-SVM and GRU

Extreme climate in alpine regions leads to frequent failures of drop-out fuses, but existing methods have low detection accuracy and response speed. This article combines MKL-SVM (Multiple Kernel Learning-Support Vector Machine) and GRU (Gated Recurrent Unit) model to achieve precise fault diagnosis of drop-out fuses, leveraging their efficient processing of high-dimensional time series data. Sensors are used to collect data such as voltage, current, temperature, and wind speed. Based on GRU, global trends, temporal hidden states, and dynamic nonlinear features of the data are extracted as inputs for subsequent MKL-SVM. In MKL-SVM, a linear kernel function is used to simplify the linear relationship between static features; targeting temporal features, RBF (Radial Basis Function) kernel effectively captures dynamic trends, fluctuations, and periodic changes in temporal features. To further enhance the classification ability of the model, the kernel function parameters, including linear kernel weights and RBF kernel widths, are adjusted through cross-validation. Finally, based on the model diagnosis results and predicted data, an optimized operation and maintenance plan is formulated to reduce maintenance costs and improve equipment operating efficiency. The results show that the fusion model can significantly improve the reliability of fault detection and adapt to the unique environmental challenges of alpine regions. The accuracy of fault diagnosis for drop-out fuses is 95%, with an MSE (mean square error) value of about 0.15 and a maintenance cost reduction of about 21.89%. This method can achieve accurate fault diagnosis and efficient optimization of operation and maintenance under extreme weather conditions.

THE IMPLEMENTATION OF GEOGRAPHICALLY WEIGHTED REGRESSION (GWR) METHOD ON OPEN UNEMPLOYMENT RATE IN REGENCY/CITY OF SUMATRA ISLAND

Unemployment is a condition where a person who is included in the labor force but does not have a job and is not actively looking for work. The number of unemployed is measured using the Open Unemployment Rate (OUR) indicator. OUR is obtained by comparing the number of job seekers and the number of labor force. This study aims to obtain a model of OUR in each district / city of Sumatra Island and what factors influence it using the Geographically Weighted Regression (GWR) method and Fixed Gaussian Kernel Function weighting, and describe predictor variables on thematic maps. The GWR method is one of the statistical methods that can prevent the presence of spatial aspects in the data. The parameters estimated by the local regression model vary at each location point and are estimated using the Weighted Least Square (WLS) method. Based on the research results obtained from this study, the GWR models obtained amounted to 154 different local models in each district / city on the island of Sumatra. Variables Labor Force Participation Rate, Population Growth Rate, Population Density and Average Years of Schooling have a significant influence on each location, meanwhile variable Percentage of Poor Population and variable Poverty Line have no influence on any location. These variables are able to explain the OUR by 57.2%, where the remaining 42.8% is explained by other factors that are not explained in the model.

Comparison of Machine Learning Algorithms in Classifying Districts/Cities in Indonesia According to the Human Development Index (HDI) in 2021

The human development index (HDI) is one of the measuring tools for achieving the quality of life of a region or even a country, including Indonesia. There are 3 basic components of the HDI, namely the dimensions of health, knowledge, and decent living. Development in Indonesia is uneven as indicated by the Human Development Index (HDI) of districts/cities in 2021 which varies greatly. The purpose of this study is to compare several machine learning algorithms to classify districts/cities in Indonesia according to the Human Development Index (HDI) in 2021. There are six machine learning algorithms used in this study, namely Artificial Neural Network (ANN), Support Vector Machine (SVM), K-Nearset Neighbor (K-NN), Random Forest, Decision Tree, and Naive Bayes. The k-Fold Cross Validation method is applied to form the training set and testing set, with 10 folds and 1 repetition. The results of the study showed that the classification results of the SVM algorithm using the Radial Basis Function (RBF) kernel parameters with sigma = 0.4864648 and C = 1 were the best among the other five algorithms with an average accuracy of 76.08% and a maximum accuracy of 88.24%.

MSBKA: A Multi-Strategy Improved Black-Winged Kite Algorithm for Feature Selection of Natural Disaster Tweets Classification.

With the advancement of the Internet, social media platforms have gradually become powerful in spreading crisis-related content. Identifying informative tweets associated with natural disasters is beneficial for the rescue operation. When faced with massive text data, choosing the pivotal features, reducing the calculation expense, and increasing the model classification performance is a significant challenge. Therefore, this study proposes a multi-strategy improved black-winged kite algorithm (MSBKA) for feature selection of natural disaster tweets classification based on the wrapper method's principle. Firstly, BKA is improved by utilizing the enhanced Circle mapping, integrating the hierarchical reverse learning, and introducing the Nelder-Mead method. Then, MSBKA is combined with the excellent classifier SVM (RBF kernel function) to construct a hybrid model. Finally, the MSBKA-SVM model performs feature selection and tweet classification tasks. The empirical analysis of the data from four natural disasters shows that the proposed model has achieved an accuracy of 0.8822. Compared with GA, PSO, SSA, and BKA, the accuracy is increased by 4.34%, 2.13%, 2.94%, and 6.35%, respectively. This research proves that the MSBKA-SVM model can play a supporting role in reducing disaster risk.

Support Vector Machine-Based Prediction Model for Healthcare Workforce Transition Success Under Decentralization

Objective To develop predictive models for healthcare workforce transition success under decentralization using Support Vector Machine (SVM) analysis and identify key determinants across organizational support domains. Methods A cross-sectional study was conducted among healthcare personnel (n=430) who transferred from Ministry of Public Health facilities to Provincial Administrative Organizations in Thailand during 2023-2024. The SVM model evaluated 37 predictor variables spanning demographic characteristics, benefits, and welfare domains. Four kernel functions were compared to identify optimal model performance, and feature importance analysis was conducted to determine key predictors. Results The linear kernel demonstrated superior performance (accuracy: 71.43%, sensitivity: 49.02%, specificity: 85.37%) compared to other kernel functions. Analysis revealed five key predictors (feature weights >0.25): competitive compensation (0.427), career development opportunities (0.358), fair promotion processes (0.336), hazardous work compensation (0.285), and educational leave opportunities (0.252). While employee qualifications (0.236) emerged as a significant demographic predictor, organizational support factors, particularly financial incentives and professional development opportunities, showed stronger predictive power for transition success. Conclusions This study represents the first application of machine learning techniques to predict healthcare personnel transition success in decentralization contexts. The SVM model effectively identified critical factors influencing workforce transitions, emphasizing the importance of balanced organizational support mechanisms. These findings provide evidence-based guidance for healthcare administrators implementing decentralization policies, offering generalizable insights for workforce management during health system reforms.

Long-term Monte Carlo-based neutrino-radiation hydrodynamics simulations for a black hole-torus system

We present our new general relativistic Monte Carlo (MC)-based neutrino radiation hydrodynamics code designed to solve axisymmetric systems with several improvements. The main improvements are as follows: (i) the development of an extended version of the implicit MC method for multispecies radiation fields; (ii) modeling of neutrino pair process rates based on a new numerically efficient and asymptotically correct fitting function for the kernel function; (iii) the implementation of new numerical limiters on the radiation-matter interaction to ensure a stable and physically correct evolution of the system. We apply our code to a black hole (BH)-torus system with a BH mass of 3M⊙, BH dimensionless spin of 0.8, and a torus mass of 0.1M⊙, which mimics a postmerger remnant of a binary neutron star merger in the case that the massive neutron star collapses to a BH within a short timescale (∼10 ms). We follow the evolution of the BH-torus system up to more than 1 s with our MC-based radiation viscous-hydrodynamics code that dynamically takes into account nonthermal pair annihilation. We find that the system evolution and the various key quantities, such as neutrino luminosity, ejecta mass, torus Ye, and pair annihilation luminosity, are broadly in agreement with the results of the previous studies. We also find that the νeν¯e pair annihilation can launch a relativistic outflow for a timescale of ∼0.1 s, and it can be energetic enough to explain some of short-hard gamma-ray bursts and the precursors. Finally, we calculate the indicators of the fast flavor instability directly from the obtained neutrino distribution functions, which indicate that the instability can occur particularly near the equatorial region of the torus. Published by the American Physical Society 2025

Smoothing Estimation of Parameters in Censored Quantile Linear Regression Model

In this paper, we propose a smoothing estimation method for censored quantile regression models. The method associates the convolutional smoothing estimation with the loss function, which is quadratically derivable and globally convex by using a non-negative kernel function. Thus, the parameters of the regression model can be computed by using the gradient-based iterative algorithm. We demonstrate the convergence speed and asymptotic properties of the smoothing estimation for large samples in high dimensions. Numerical simulations show that the smoothing estimation method for censored quantile regression models improves the estimation accuracy, computational speed, and robustness over the classical parameter estimation method. The simulation results also show that the parametric methods perform better than the KM method in estimating the distribution function of the censored variables. Even if there is an error setting in the distribution estimation, the smoothing estimation does not fluctuate too much.

The Spectral Envelope Estimate via Dirichlet Kernel Function

In this article the spectral envelope function for a nonstationary time series is estimated by using the Dirichlet kernel function. we divide the time series into a number of segmentations, then the nonstationary problem is solved to reduce its effect. The simulation and the sequence of the nitrogenous bases of the DNA of Escherichia coli bacteria are used to implement the goal of this article.