Article
Automation & Control Systems
Weidong Xie, Yushan Fang, Kun Yu, Xin Min, Wei Li
Summary: MFRAG is a new hybrid feature selection method that mimics the natural principle of survival of the fittest by enhancing the stability and reliability of the selection process through fusion mechanisms and integrated models, and guides the evolutionary process through a set of lists generated by a feature fusion model.
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yu Xue, Haokai Zhu, Jiayu Liang, Adam Slowik
Summary: Feature selection is a crucial pre-processing technique for classification, aiming to enhance classification accuracy by removing irrelevant or redundant features. This study introduces a multi-objective genetic algorithm with an adaptive operator selection mechanism, which effectively addresses high-dimensional feature selection problems.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Jianbin Ma, Xiaoying Gao, Ying Li
Summary: The purpose of feature construction is to create new high level features from the original features. Genetic Programming (GP) tends to overfit the training set and generalize poorly with the deepening of evolution in wrapper-based feature construction, especially when the sample size is small. Overfitting has been widely studied in classification models but not commonly explored in feature construction.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Guopeng Liu, Jianbin Ma, Tongle Hu, Xiaoying Gao
Summary: This paper proposes a feature selection method using genetic programming and feature ranking, which can achieve better classification performance by further reducing the number of selected features using a multi-criteria fitness function.
CONNECTION SCIENCE
(2022)
Article
Computer Science, Interdisciplinary Applications
R. J. Kuo, Muhammad Rakhmat Setiawan, Thi Phuong Quyen Nguyen
Summary: This study introduces a novel data analytics-based sequential clustering and classification (SCC) approach, named deep MOSCA-SCC, which integrates multi-objective sine-cosine algorithm (MOSCA), deep clustering technique, and classification algorithms. The method shows better performance in terms of clustering sum of squared error and classification accuracy compared to other benchmark algorithms.
COMPUTERS & INDUSTRIAL ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Motahare Akhavan, Seyed Mohammad Hossein Hasheminejad
Summary: A new two-phase gene selection method for microarray data is proposed in this study. This method reduces the number of genes significantly and improves the classification accuracy through anomaly detection and guided genetic algorithm.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Engineering, Civil
Javier Echanobe, Koldo Basterretxea, Ines del Campo, Victoria Martinez, Naiara Vidal
Summary: Driver Assistance Systems (DAS) have laid the groundwork for autonomous vehicles, and reducing system complexity is a valuable contribution in this field.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Farrukh Hasan Syed, Muhammad Atif Tahir, Muhammad Rafi, Mir Danish Shahab
Summary: Multi-target regression is an emerging area in machine learning focusing on predicting the values of multiple target variables. These problems are common in real life scenarios and have attracted increasing interest and research in recent years. Combining genetic algorithms with semi-supervised techniques can effectively address the challenges in this field.
APPLIED INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Yu Zhou, Wenjun Zha, Junhao Kang, Xiao Zhang, Xu Wang
Summary: This paper proposes a problem-specific non-dominated sorting genetic algorithm (PS-NSGA) that can minimize three objectives of feature selection. By applying an accuracy-preferred domination operator and a quick bit mutation, the algorithm converges faster and better, achieving competitive classification accuracy in experiments.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Farrukh Hasan Syed, Muhammad Atif Tahir, Jaroslav Frnda, Muhammad Rafi, Muhammad Shahid Anwar, Jan Nedoma
Summary: Multi Target Regression (MTR) is a machine learning method for predicting multiple real-valued outputs simultaneously. This research proposes multiple feature subset alternatives for MTR using genetic algorithm and compares their performance with MTR algorithms. Experimental results indicate that optimal and structured feature selection can significantly improve performance and yield comparatively simple MTR models.
Article
Computer Science, Information Systems
Xing Yong Kek, Cheng Siong Chin, Ye Li
Summary: In this paper, a genetic algorithm is applied for feature selection on wavelet scattering coefficients, reducing the frequency dimension and improving classification performance. Multiple timescales and feature selection are explored, achieving better accuracy.
Article
Computer Science, Artificial Intelligence
Che Xu, Shuwen Zhang
Summary: This paper proposes a framework for sequential instance selection based on the Genetic Algorithm to address the balance between individual accuracy and diversity in ensemble models. The framework overcomes the limitations of the Genetic Algorithm in high-dimensional tasks and provides a way to balance accuracy and diversity by searching appropriate training data subsets. The predictions of the component classifiers are combined using a weighted majority voting rule.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Computer Science, Interdisciplinary Applications
Xiongshi Deng, Min Li, Shaobo Deng, Lei Wang
Summary: This paper proposes a two-stage gene selection method combining XGBoost and XGBoost-MOGA for cancer classification in microarray datasets. The experimental results show that this method outperforms other state-of-the-art algorithms in terms of accuracy, F-score, precision, and recall.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Rui Zhang, Zuoquan Zhang, Di Wang, Marui Du
Summary: HSMOGA is a novel feature selection algorithm that introduces a hybrid filter, Symmetrical Complementary Coefficient, and a new method to limit feature subset size. It uses a Pareto-based ranking function to solve multi-objective problems and precalculates knowledge using a step called knowledge reserve, leading to faster convergence of solutions. Experimental results show that HSMOGA outperforms other nine feature selection algorithms in terms of performance metrics such as kappa coefficient, accuracy, and G-mean.
APPLIED INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
K. Aditya Shastry, H. A. Sanjay
Summary: Data pre-processing is a technique that transforms raw data into a useful format for machine learning, with feature selection and feature extraction being significant components. This study proposes a hybrid strategy using modified Genetic Algorithm and weighted Principal Component Analysis for selecting and extracting features from agricultural datasets, resulting in significant improvements in benchmark and real-world farming datasets.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Chemistry, Analytical
Pan Huang, Xiaoheng Tan, Chen Chen, Xiaoyi Lv, Yongming Li
Summary: This study improved the accuracy and generalization ability of cervical cancer WSI recognition by designing an image processing algorithm and serially fusing deep features, showing superior performance especially in LSIL recognition area.
Article
Critical Care Medicine
Meifang Yin, Jiangfeng Li, Lixian Huang, Yongming Li, Mingzhou Yuan, Yongquan Luo, Ubaldo Armato, Lijun Zhang, Yating Wei, Yuanyuan Li, Jiawen Deng, Pin Wang, Jun Wu
Summary: A novel technique using NIRS and SVM was developed to rapidly identify microbial species and drug-resistant bacteria, and successfully diagnose colonization and infection models in pig wounds. The technique demonstrated high accuracy and rapid identification, showing potential for the rapid diagnosis of infected wounds.
Article
Clinical Neurology
Qian Yu, Xiaoya Zou, Fengying Quan, Zhaoying Dong, Huimei Yin, Jinjing Liu, Hongzhou Zuo, Jiaman Xu, Yu Han, Dezhi Zou, Yongming Li, Oumei Cheng
Summary: This study compared the acoustic parameters of Parkinson's disease patients with and without freezing of gait (FOG), and explored the ability of voice features to distinguish between the two groups. The results showed that patients with FOG had more severe voice impairment during the ON state.
JOURNAL OF NEURAL TRANSMISSION
(2022)
Article
Engineering, Biomedical
Pin Wang, Pufei Li, Yongming Li, Jin Xu, Mingfeng Jiang
Summary: This study proposes a deep transferred semi-supervised domain adaptation model for classification of histopathological whole slide images (WSIs). By utilizing a transferred pre-trained network and semi-supervised domain adaptation, accurate classification results can be achieved with limited labeled samples.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Review
Radiology, Nuclear Medicine & Medical Imaging
Yang Li, Qiannan Shen, Mingfeng Jiang, Lingyan Zhu, Yongming Li, Pin Wang, Tie-Qiang Li
Summary: This paper proposes a low-rank plus sparse decomposition approach for reconstructing dynamic magnetic resonance imaging (dMRI). The optimization problem is solved by using a sequentially truncated higher-order singular value decomposition method and an iterative soft-thresholding algorithm. Experimental results demonstrate that the proposed method achieves better performance in terms of reconstruction speed and quality.
CURRENT MEDICAL IMAGING
(2022)
Article
Engineering, Electrical & Electronic
Pin Wang, Pufei Li, Yongming Li, Jin Xu, Fang Yan, Mingfeng Jiang
Summary: In this study, a breast histopathology image classification method based on deep manifold fusion of multilayer features, LPMF2Net, is proposed. By fusing features at different levels and applying local preserving projection and adaptive adjustment of the projection matrix, the model's effectiveness is verified through experimental results.
DIGITAL SIGNAL PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Fan Li, Xiaoheng Zhang, Pin Wang, Yongming Li
Summary: Imbalanced learning is a significant and challenging problem in the fields of machine learning and data mining. Traditional clustering methods have limitations in handling imbalanced datasets, thus a new algorithm based on deep instance envelope network and minimum interlayer discrepancy mechanism is proposed to address these issues.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Yongming Li, Chengyu Liu, Pin Wang, Hehua Zhang, Anhai Wei, Yanling Zhang
Summary: In recent years, the study of machine learning-based speech recognition methods for Parkinson's disease has gained popularity. However, the focus has been more on feature learning and classifier design rather than sample optimization. This paper proposes a multi-type transformation ensemble algorithm for PD speech samples based on a subject envelope (MTEA) to address this issue. The algorithm performs multi-type transformations on the segment samples within a subject envelope, improving the quality of sample transformation through a joint structure consistency mechanism (JSCM) and fusing the results of multiple classifiers using a sparse weighted fusion mechanism. Experimental results demonstrate the effectiveness of the proposed method in improving classification accuracy.
APPLIED INTELLIGENCE
(2023)
Article
Biology
Zhengyang Wu, Guifeng Xia, Xiaoheng Zhang, Fayuan Zhou, Jing Ling, Xin Ni, Yongming Li
Summary: A novel 3D lumbar vertebrae location and segmentation method based on the fusion envelope of 2D hybrid visual projection images is proposed in this paper, which can better achieve real-time and accurate lumbar segmentation compared to existing methods.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Engineering, Biomedical
Yongming Li, Jin Xu, Pin Wang, Pufei Li, Gongxin Yang, Rui Chen
Summary: A manifold reconstructed semi-supervised domain adaptation model is proposed for whole slide images' classification. A transferred network Bre-Net is used to extract features from multiple layers of source and target domains, and the fused features are aligned to characterize the cell structure of the target domain. A novel manifold reconstructed domain adaptation method is used to obtain the low-dimensional embedding of the fused features and minimize the cross-domain discrepancy. Patch-level prediction probabilities are aggregated for final image-level classification.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Energy & Fuels
Hao Chen, Qifeng Liu, Yongming Li, Chen Huang, Huaiqing Zhang, Yinxiang Xu
Summary: This article proposes a method of near-field measurement and a modeling method of powerful electromagnetic equipment radiation based on field distribution characteristics. The near-field measurement data are obtained by sparse and uniform sampling, and the measurement plane is separated into several regions based on magnetic field distribution characteristics. The required near-field measurement data for modeling are obtained by further sampling in the region with a large magnetic field amplitude. Finally, the equivalent radiation model is obtained by the equivalent dipole method. Simulation and experiments show that the method can significantly reduce the amount of measurement data and testing time while improving the modeling efficiency and accuracy.
Article
Chemistry, Multidisciplinary
Yi Zhang, Jie Ma, Xiaolin Qin, Yongming Li, Zuwei Zhang
Summary: This study proposes a machine-learning-based method using wearable sensors for early diagnosis, addressing the challenge of obtaining high-quality and large amounts of data. The proposed algorithm achieves significantly better results compared to existing algorithms, improving diagnostic accuracy on multiple criteria.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Fan Li, Bo Wang, Yinghua Shen, Pin Wang, Yongming Li
Summary: This paper proposes an imbalanced ensemble learning algorithm based on weighted projection clustering grouping and consistent fuzzy sample transformation. It utilizes a weighted projection clustering combination framework to obtain high-quality clusters and applies a stage-wise hybrid sampling algorithm for de-overlapping and balancing of subsets. Additionally, a local-global structure consistency mechanism is constructed to improve the quality of samples in subsets. Experimental results demonstrate the superiority of the proposed algorithm in terms of anti-overlapping, Recall, F1-M, G-M, AUC, and diversity.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Interdisciplinary Applications
Yiwen Wang, Fan Li, Xiaoheng Zhang, Pin Wang, Yongming Li, Yanling Zhang
Summary: Machine learning-based Parkinson's disease (PD) speech diagnosis is a current research hotspot. This study proposes a novel algorithm using sample compression and ensemble learning mechanisms to extract stable diagnostic markers and meet the requirements of clinical applications.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Article
Engineering, Biomedical
Pin Wang, Gongxin Yang, Yongming Li, Pufei Li, Yurou Guo, Rui Chen
Summary: This paper proposes a deep sample clustering unsupervised domain adaptation method based on deep fusion feature for breast histopathology image classification. The method accurately classifies unlabeled data by combining clustering and classification, and has the potential for clinical application.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2024)