Xiaoyong Liu
Xiaoyong Liu has a total of 35 co-authors. Here are the top co-authors by number of shared publications:
- Chengbin Zeng
School of Automation Engineering, Moutai Institute, Renhuai 564507, China
3 shared publications - Fabin Zhang
School of Automation Engineering, Moutai Institute, Renhuai 564507, China
2 shared publications - Genglong Yan
School of Automation Engineering, Moutai Institute, Renhuai 564507, China
2 shared publications - Zhonghua Yun
School of Instrument Science and Engineering, Southeast University, Nanjing, China
2 shared publications - Aijia Ouyang
Zunyi Normal University
1 shared publication - Ashish Kumar Luhach
Department of Electrical and Communication Engineering , The Papua New Guinea University of Technology , Lae , Papua New Guinea , unitech.ac.pg
1 shared publication
Latest Publications
- Open Access
- 18 February 2026
- Published by MDPI AG in Algorithms
- Vol. 19 (2) , 160
A Sparse -Norm Regularized Least Squares Support Vector RegressionAlthough Least Squares Support Vector Regression (LSSVR) reduces the hyperparameter space to two, it sacrifices sparsity, causing all training samples to become support vectors and increasing storage costs. In contrast, standard Support Vector Regression (SVR) preserves sparsity but requires tuning three highly coupled hyperparameters, leading to higher computational burden. To address these limitations, this paper proposes a sparse -norm regularized least squares SVR framework that incorporates the infinity norm of approximation errors into both the objective function and inequality constraints. The resulting optimization problem minimizes model complexity while controlling the maximum prediction deviation through a single slack variable, thereby transforming the conventional three-hyperparameter SVR tuning task into a two-parameter problem involving only the regularization coefficient and kernel width. This formulation restores sparsity by enabling a compact support vector set, while preserving the stability and convexity advantages of LSSVR. Experiments on both static and dynamic datasets demonstrate that the proposed method consistently achieves higher predictive accuracy and improved robustness compared with standard SVR and LSSVR. These results indicate that the proposed -norm regularized framework offers a mathematically principled and computationally efficient alternative for sparse, robust, and scalable regression modeling.- 18 February 2026
- Published by MDPI AG in Algorithms
- Vol. 19 (2) , 160
Open AccessA Sparse -Norm Regularized Least Squares Support Vector RegressionAlthough Least Squares Support Vector Regression (LSSVR) reduces the hyperparameter space to two, it sacrifices sparsity, causing all training samples to become support vectors and increasing storage costs. In contrast, standard Support Vector Regression (SVR) preserves sparsity but requires tuning three highly coupled hyperparameters, leading to higher computational burden. To address these limitations, this paper proposes a sparse -norm regularized least squares SVR framework that incorporates the infinity norm of approximation errors into both the objective function and inequality constraints. The resulting optimization problem minimizes model complexity while controlling the maximum prediction deviation through a single slack variable, thereby transforming the conventional three-hyperparameter SVR tuning task into a two-parameter problem involving only the regularization coefficient and kernel width. This formulation restores sparsity by enabling a compact support vector set, while preserving the stability and convexity advantages of LSSVR. Experiments on both static and dynamic datasets demonstrate that the proposed method consistently achieves higher predictive accuracy and improved robustness compared with standard SVR and LSSVR. These results indicate that the proposed -norm regularized framework offers a mathematically principled and computationally efficient alternative for sparse, robust, and scalable regression modeling. - Open Access
A Dual-Norm Support Vector Machine: Integrating L1 and L∞ Slack Penalties for Robust and Sparse Classification
This paper presents a novel support vector machine (SVM) classification approach that simultaneously accounts for both overall and extreme misclassification errors via a dual-norm regularization strategy. Traditional SVMs minimize the -norm of slack variables to control global misclassification, while least squares SVM (LSSVM) minimizes the sum of squared errors. In contrast, our method preserves the classical -norm penalty to maintain overall classification fidelity and incorporates an additional -norm term to penalize the largest slack variable, thereby constraining the worst-case margin violation. This composite objective yields a more robust and generalizable classifier, particularly effective when occasional large deviations disproportionately affect decision boundaries. The resulting optimization problem minimizes a regularized objective combining the model norm, the sum of slack variables, and the maximum slack variable, with two hyperparameters, and , balancing global error against extremal robustness. By formulating the problem under convex constraints, the optimization remains tractable and guarantees a globally optimal solution. Experimental evaluations on benchmark datasets demonstrate that the proposed method achieves comparable or superior classification accuracy while reducing the impact of outliers and maintaining a sparse model structure. These results underscore the advantage of jointly enforcing and penalties, providing an effective mechanism to balance average performance with worst-case error sensitivity in support vector classification.Open AccessA Dual-Norm Support Vector Machine: Integrating L1 and L∞ Slack Penalties for Robust and Sparse Classification
This paper presents a novel support vector machine (SVM) classification approach that simultaneously accounts for both overall and extreme misclassification errors via a dual-norm regularization strategy. Traditional SVMs minimize the -norm of slack variables to control global misclassification, while least squares SVM (LSSVM) minimizes the sum of squared errors. In contrast, our method preserves the classical -norm penalty to maintain overall classification fidelity and incorporates an additional -norm term to penalize the largest slack variable, thereby constraining the worst-case margin violation. This composite objective yields a more robust and generalizable classifier, particularly effective when occasional large deviations disproportionately affect decision boundaries. The resulting optimization problem minimizes a regularized objective combining the model norm, the sum of slack variables, and the maximum slack variable, with two hyperparameters, and , balancing global error against extremal robustness. By formulating the problem under convex constraints, the optimization remains tractable and guarantees a globally optimal solution. Experimental evaluations on benchmark datasets demonstrate that the proposed method achieves comparable or superior classification accuracy while reducing the impact of outliers and maintaining a sparse model structure. These results underscore the advantage of jointly enforcing and penalties, providing an effective mechanism to balance average performance with worst-case error sensitivity in support vector classification. - Open Access
Linear Programming-Based Sparse Kernel Regression with L1-Norm Minimization for Nonlinear System Modeling
This paper integrates -norm structural risk minimization with -norm approximation error to develop a new optimization framework for solving the parameters of sparse kernel regression models, addressing the challenges posed by complex model structures, over-fitting, and limited modeling accuracy in traditional nonlinear system modeling. The first -norm regulates the complexity of the model structure to maintain its sparsity, while another -norm is essential for ensuring modeling accuracy. In the optimization of support vector regression (SVR), the -norm structural risk is converted to an -norm framework through the condition of non-negative Lagrange multipliers. Furthermore, -norm optimization for modeling accuracy is attained by minimizing the maximum approximation error. The integrated -norm of structural risk and approximation errors creates a new, simplified optimization problem that is solved using linear programming (LP) instead of the more complex quadratic programming (QP). The proposed sparse kernel regression model has the following notable features: (1) it is solved through relatively simple LP; (2) it effectively balances the trade-off between model complexity and modeling accuracy; and (3) the solution is globally optimal rather than just locally optimal. In our three experiments, the sparsity metrics of were 2.67%, 1.40%, and 0.8%, with test RMSE values of 0.0667, 0.0701, 0.0614 (sinusoidal signal), and 0.0431 (step signal), respectively. This demonstrates the balance between sparsity and modeling accuracy.Open AccessLinear Programming-Based Sparse Kernel Regression with L1-Norm Minimization for Nonlinear System Modeling
This paper integrates -norm structural risk minimization with -norm approximation error to develop a new optimization framework for solving the parameters of sparse kernel regression models, addressing the challenges posed by complex model structures, over-fitting, and limited modeling accuracy in traditional nonlinear system modeling. The first -norm regulates the complexity of the model structure to maintain its sparsity, while another -norm is essential for ensuring modeling accuracy. In the optimization of support vector regression (SVR), the -norm structural risk is converted to an -norm framework through the condition of non-negative Lagrange multipliers. Furthermore, -norm optimization for modeling accuracy is attained by minimizing the maximum approximation error. The integrated -norm of structural risk and approximation errors creates a new, simplified optimization problem that is solved using linear programming (LP) instead of the more complex quadratic programming (QP). The proposed sparse kernel regression model has the following notable features: (1) it is solved through relatively simple LP; (2) it effectively balances the trade-off between model complexity and modeling accuracy; and (3) the solution is globally optimal rather than just locally optimal. In our three experiments, the sparsity metrics of were 2.67%, 1.40%, and 0.8%, with test RMSE values of 0.0667, 0.0701, 0.0614 (sinusoidal signal), and 0.0431 (step signal), respectively. This demonstrates the balance between sparsity and modeling accuracy. - Open Access
- 14 October 2022
- Published by Frontiers Media SA in Frontiers in Genetics
- Vol. 13, 1032779
Drug-protein interaction prediction via variational autoencoders and attention mechanisms
During the process of drug discovery, exploring drug-protein interactions (DPIs) is a key step. With the rapid development of biological data, computer-aided methods are much faster than biological experiments. Deep learning methods have become popular and are mainly used to extract the characteristics of drugs and proteins for further DPIs prediction. Since the prediction of DPIs through machine learning cannot fully extract effective features, in our work, we propose a deep learning framework that uses variational autoencoders and attention mechanisms; it utilizes convolutional neural networks (CNNs) to obtain local features and attention mechanisms to obtain important information about drugs and proteins, which is very important for predicting DPIs. Compared with some machine learning methods on the C.elegans and human datasets, our approach provides a better effect. On the BindingDB dataset, its accuracy (ACC) and area under the curve (AUC) reach 0.862 and 0.913, respectively. To verify the robustness of the model, multiclass classification tasks are performed on Davis and KIBA datasets, and the ACC values reach 0.850 and 0.841, respectively, thus further demonstrating the effectiveness of the model.
- 14 October 2022
- Published by Frontiers Media SA in Frontiers in Genetics
- Vol. 13, 1032779
Open AccessDrug-protein interaction prediction via variational autoencoders and attention mechanisms
During the process of drug discovery, exploring drug-protein interactions (DPIs) is a key step. With the rapid development of biological data, computer-aided methods are much faster than biological experiments. Deep learning methods have become popular and are mainly used to extract the characteristics of drugs and proteins for further DPIs prediction. Since the prediction of DPIs through machine learning cannot fully extract effective features, in our work, we propose a deep learning framework that uses variational autoencoders and attention mechanisms; it utilizes convolutional neural networks (CNNs) to obtain local features and attention mechanisms to obtain important information about drugs and proteins, which is very important for predicting DPIs. Compared with some machine learning methods on the C.elegans and human datasets, our approach provides a better effect. On the BindingDB dataset, its accuracy (ACC) and area under the curve (AUC) reach 0.862 and 0.913, respectively. To verify the robustness of the model, multiclass classification tasks are performed on Davis and KIBA datasets, and the ACC values reach 0.850 and 0.841, respectively, thus further demonstrating the effectiveness of the model.
- 1 January 2022
- Published by Four Dimensions Publishing Group Inc. in Education Theory: Teaching and Learning
- p. 113-117
Research on the Reform of "Three Education" in Higher Vocational Colleges Under the Promotion of "Double High" Construction
Implementing the national vocational education reform implementation plan to build high-level vocational schools with Chinese characteristics and specialty construction plan is an important measure to cultivate high-quality technical and skilled talents. The reform of "three educations", namely the reform of teachers, teaching materials and teaching methods, is the key link in the implementation of the construction of "double highs". Fully understanding the significance of the "three education" reform and grasping the implementation path will help to cultivate high-quality technical talents and promote the high-quality development of higher vocational colleges.
- 1 January 2022
- Published by Four Dimensions Publishing Group Inc. in Education Theory: Teaching and Learning
- p. 113-117
Research on the Reform of "Three Education" in Higher Vocational Colleges Under the Promotion of "Double High" Construction
Implementing the national vocational education reform implementation plan to build high-level vocational schools with Chinese characteristics and specialty construction plan is an important measure to cultivate high-quality technical and skilled talents. The reform of "three educations", namely the reform of teachers, teaching materials and teaching methods, is the key link in the implementation of the construction of "double highs". Fully understanding the significance of the "three education" reform and grasping the implementation path will help to cultivate high-quality technical talents and promote the high-quality development of higher vocational colleges.
- 1 October 2021
- Published by Elsevier in Sustainable Energy Technologies and Assessments
- Vol. 47, 101477
The optimized energy-efficient sensible edge processing model for the internet of vehicles in smart cities
- 1 October 2021
- Published by Elsevier in Sustainable Energy Technologies and Assessments
- Vol. 47, 101477
The optimized energy-efficient sensible edge processing model for the internet of vehicles in smart cities
- Open Access
DaSGD: Squeezing SGD Parallelization Performance in Distributed Training Using Delayed Averaging
The state-of-the-art deep learning algorithms rely on distributed training systems to tackle the increasing sizes of models and training data sets. Minibatch stochastic gradient descent (SGD) algorithm requires workers to halt forward/back propagations, to wait for gradients aggregated from all workers, and to receive weight updates before the next batch of tasks. This synchronous execution model exposes the overheads of gradient/weight communication among a large number of workers in a distributed training system. We propose a new SGD algorithm, DaSGD (Local SGD with Delayed Averaging), which parallelizes SGD and forward/back propagations to hide 100% of the communication overhead. By adjusting the gradient update scheme, this algorithm uses hardware resources more efficiently and reduces the reliance on the low-latency and high-throughput inter-connects. The theoretical analysis and the experimental results show its convergence rate O(1/sqrt(K)), the same as SGD. The performance evaluation demonstrates it enables a linear performance scale-up with the cluster size.
Open AccessDaSGD: Squeezing SGD Parallelization Performance in Distributed Training Using Delayed Averaging
The state-of-the-art deep learning algorithms rely on distributed training systems to tackle the increasing sizes of models and training data sets. Minibatch stochastic gradient descent (SGD) algorithm requires workers to halt forward/back propagations, to wait for gradients aggregated from all workers, and to receive weight updates before the next batch of tasks. This synchronous execution model exposes the overheads of gradient/weight communication among a large number of workers in a distributed training system. We propose a new SGD algorithm, DaSGD (Local SGD with Delayed Averaging), which parallelizes SGD and forward/back propagations to hide 100% of the communication overhead. By adjusting the gradient update scheme, this algorithm uses hardware resources more efficiently and reduces the reliance on the low-latency and high-throughput inter-connects. The theoretical analysis and the experimental results show its convergence rate O(1/sqrt(K)), the same as SGD. The performance evaluation demonstrates it enables a linear performance scale-up with the cluster size.
- 19 August 2019
- Published by SAGE Publications in Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
- Vol. 37 (3) , 3991-4001
Fault detection based on LP-SVR interval regression model with L 1 -Norm minimization
In this paper, a novel method for fault detection based on an adaptive interval regression model characterized by the upper regression model (URM) and lower regression model (LRM) has been proposed. Applying the proposed method, a confidence band for the measured data, derived in the normal operating conditions of a system, is constructed.The method combines the superiorities of model sparse representation and computational efficiency of linear programming support vector regression (LP-SVR) with some ideas from L 1 -norm on approximation errors. First, the upper and lower L 1 -norms with respect to upper bound approximation error are considered, and the both norms subject to respective constraints are integrated into LP-SVR to form new upper and lower optimization problems, respectively. Following that, optimization problem corresponding to URM and LRM are solved by linear programming and interval regression model is thus constructed to judge whether the fault occurs or not. The proposed method returns an interval output as opposed to a point output. Finally, the efficacy of this method is demonstrated by applying it on the benchmark Tennessee Eastman problem, and has been compared with conventional techniques such as principal component analysis (PCA), dynamic-PCA (DPCA) and One-Class Support Vector Machine(1-class SVM). It is shown that the proposed method is superior to those approaches in terms of performance measure of detection latency.
- 19 August 2019
- Published by SAGE Publications in Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology
- Vol. 37 (3) , 3991-4001
Fault detection based on LP-SVR interval regression model with L 1 -Norm minimization
In this paper, a novel method for fault detection based on an adaptive interval regression model characterized by the upper regression model (URM) and lower regression model (LRM) has been proposed. Applying the proposed method, a confidence band for the measured data, derived in the normal operating conditions of a system, is constructed.The method combines the superiorities of model sparse representation and computational efficiency of linear programming support vector regression (LP-SVR) with some ideas from L 1 -norm on approximation errors. First, the upper and lower L 1 -norms with respect to upper bound approximation error are considered, and the both norms subject to respective constraints are integrated into LP-SVR to form new upper and lower optimization problems, respectively. Following that, optimization problem corresponding to URM and LRM are solved by linear programming and interval regression model is thus constructed to judge whether the fault occurs or not. The proposed method returns an interval output as opposed to a point output. Finally, the efficacy of this method is demonstrated by applying it on the benchmark Tennessee Eastman problem, and has been compared with conventional techniques such as principal component analysis (PCA), dynamic-PCA (DPCA) and One-Class Support Vector Machine(1-class SVM). It is shown that the proposed method is superior to those approaches in terms of performance measure of detection latency.
- Open Access
- 10 December 2018
- Published by Wiley in Mathematical Problems in Engineering
- Vol. 2018, 1-13
Fuzzy Weighted Least Squares Support Vector Regression with Data Reduction for Nonlinear System Modeling
This paper proposes a fuzzy weighted least squares support vector regression (FW-LSSVR) with data reduction for nonlinear system modeling based only on the measured data. The proposed method combines the advantages of data reduction with some ideas of fuzzy weighted mechanism. It not only possesses the capability of illuminating local characteristic of the modeled plant but also can deal with the problem of boundary effects resulted from local LSSVR method when the modeled data is at the boundary of whole data subset. Furthermore, in comparison of the SVR, the proposed method only utilizes fewer hyperparameters to construct model, and the overlap factor can be chosen in relatively smaller value than SVR to further reduce more computational time. First of all, distilling the original input space into several regions with fuzzy partition by applying Gustafson-Kessel clustering algorithm (GKCA) is a foundation for data reduction and the overlap factor is introduced to reduce the size of subsets. Following that, those subset regression models (SRMs) which can be simultaneously solved by LSSVR are integrated into an overall output of the estimated nonlinear system by fuzzy weighted. Finally, the proposed method is demonstrated by experimental analysis and compared with local LSSVR, weighted SVR, and global LSSVR methods by using the index of computational time and root-mean-square error (RMSE).- 10 December 2018
- Published by Wiley in Mathematical Problems in Engineering
- Vol. 2018, 1-13
Open AccessFuzzy Weighted Least Squares Support Vector Regression with Data Reduction for Nonlinear System Modeling
This paper proposes a fuzzy weighted least squares support vector regression (FW-LSSVR) with data reduction for nonlinear system modeling based only on the measured data. The proposed method combines the advantages of data reduction with some ideas of fuzzy weighted mechanism. It not only possesses the capability of illuminating local characteristic of the modeled plant but also can deal with the problem of boundary effects resulted from local LSSVR method when the modeled data is at the boundary of whole data subset. Furthermore, in comparison of the SVR, the proposed method only utilizes fewer hyperparameters to construct model, and the overlap factor can be chosen in relatively smaller value than SVR to further reduce more computational time. First of all, distilling the original input space into several regions with fuzzy partition by applying Gustafson-Kessel clustering algorithm (GKCA) is a foundation for data reduction and the overlap factor is introduced to reduce the size of subsets. Following that, those subset regression models (SRMs) which can be simultaneously solved by LSSVR are integrated into an overall output of the estimated nonlinear system by fuzzy weighted. Finally, the proposed method is demonstrated by experimental analysis and compared with local LSSVR, weighted SVR, and global LSSVR methods by using the index of computational time and root-mean-square error (RMSE). - 6 August 2015
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 35 (7) , 2331-2350
Strong Tracking Filter-Based Fault Diagnosis of Networked Control System with Multiple Packet Dropouts and Parameter Perturbations
- 6 August 2015
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 35 (7) , 2331-2350
Strong Tracking Filter-Based Fault Diagnosis of Networked Control System with Multiple Packet Dropouts and Parameter Perturbations