检索结果-内蒙古大学图书馆

Electroencephalogram emotion recognition and analysis based on feature fusion and stable learning

Biomedical Signal Processing and Control 2025年 108卷

作者： Shi, Kaiting Gong, Yifan Ouyang, Rujie Yang, Lijun Yang, Xiaohui Zheng, Chen School of Mathematics and Statistics Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan University Kaifeng475004 China Center for Applied Mathematics of Henan Province Henan University Zhengzhou450046 China

Existing methods based on Graph Convolutional Networks (GCNs) have shown promise for electroencephalogram (EEG) emotion recognition. However, fully leveraging EEG electrode topology and extracting fundamental features remain areas of ongoing research. In this study, we propose a stable fusion graph convolutional network (SFGCN), an effective and high-performance classification model that integrates multiple graph topologies with stable learning. Specifically, we employ linear and Granger causality-based graph topology construction methods and introduce a squeeze-and-excitation (SE) block to enhance feature fusion, enabling a comprehensive representation of complex graph information. To address spurious correlations in feature extraction, we incorporate stable learning to ensure feature independence, thereby improving model performance. Additionally, causal brain network analysis is utilized to identify critical EEG channels, reducing computational complexity and data dimensions. Granger causality analysis reflects the information flow and functional connectivity differences in brain networks, facilitating the selection of key channels. SFGCN achieves classification accuracies of approximately 98% and 92% on the DEAP and DREAMER datasets, respectively. On DEAP, 20 critical channels were selected from 32 channels using causal brain network analysis, and their effectiveness was validated through comparative experiments. The results demonstrate that SFGCN not only improves classification accuracy but also enhances model stability. The integration of graph topology with effective feature extraction underscores the potential of this approach for practical applications in EEG-based emotion recognition. © 2025 Elsevier Ltd

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

L1/2Regularization-Based Deep Incremental Non-negative Matrix Factorization for Tumor Recognition 2021

L1/2Regularization-Based Deep Incremental Non-negative Matri...

引用

5th International Conference on Biological Information and Biomedical engineering, BIBE 2021

作者： Yan, Lulu Yang, Xiaohui Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan University China

ISBN: (纸本)9781450389297

Non-negative matrix factorization (NMF) is an effective technique for feature representation learning and dimensionality reduction. However, there are two critical challenges for improving the performance of NMF-based methods. One is the sparsity of representation, the other is the sensitivity to the initial value of the iteration, which seriously affects the performance of NMF. To solve the problems, L1/2 regularization is skillfully selected to characterize the sparsity of the data. Furthermore, a layer-wise pre-training strategy in deep learning is used to alleviate the effect of the initial value on NMF, whereby complex network structure is avoided. As such, a L1/2 regularization-based deep NMF (L1/2-DNMF) model is proposed in this study, such that a more stable and sparse deep representation is obtained. Moreover, incremental learning is introduced to reduce the high computational complexity of L1/2-DNMF model, called L1/2-DINMF model, which is suitable for online processing. Experiment results on genetic data-based tumor recognition verify that the proposed L1/2-DINMF model outperforms the classic and state-of-the-art methods. © 2021 ACM.

关键词： Non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

EEG-Based Depression Recognition Using Intrinsic Time-scale Decomposition and Temporal Convolution Network 2021

EEG-Based Depression Recognition Using Intrinsic Time-scale ...

引用

5th International Conference on Biological Information and Biomedical engineering, BIBE 2021

作者： Wang, Yixin Liu, Fengrui Yang, Lijun School of Mathematics and Statistics Henan University Kaifeng China Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan University Kaifeng China

ISBN: (纸本)9781450389297

The diagnosis and treatment of depression is very important since it brings a heavy burden to family and society. Because of the high sensitivity, relatively low cost, and convenient recording, electroencephalogram (EEG) has become an important tool for monitoring brain activity and is gradually being used in the auxiliary diagnosis of mental diseases. EEG signals are typically non-linear and non-stationary. Therefore, they are suitable to be dealt with by time-frequency analysis technique. In this paper, we propose a strategy that combines the time-frequency analysis technique and temporal convolution network for depression recognition. Firstly, we use the method of intrinsic time-scale decomposition to decompose each EEG recording to several components. And secondly, some statistical indices are calculated from the instantaneous amplitudes and instantaneous frequencies of these components to form the feature vectors. Thirdly, an improved temporal convolution network (TCN) is used to detect the depression from normal controls. Temporal convolution network is not only suitable for the sequence model, but also retains the parallel computing characteristics of the convolutional neural network. To improve the model performance, we further modify the original softmax loss of TCN as L-softmax. Experiments show the effectiveness of the proposed model. Furthermore, we find that the depressive patients and normal controls shows different patterns through functional connectivity analysis. Our analysis results can be used as an auxiliary tool to help psychiatrists diagnose patients with depression. © 2021 ACM.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A universal protocol for bidirectional controlled teleportation with network coding

引用

Communications in Theoretical Physics 2021年第10期73卷 107-113页

作者： Meng-Yao He Song-Ya Ma Kun-Peng Kang School of Mathematics and Statistics Henan UniversityKaifeng 475004China Information Security Center State Key Laboratory of Networking and Switching TechnologyBeijing University of Posts and TelecommunicationsBeijing 100876China Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan UniversityKaifeng 475004China

We investigate bidirectional teleportation that works in a fair and efficient manner. Two explicit protocols are proposed to realize bidirectional teleportation with a controller. One is a symmetric protocol for two-qubit states. The other is an asymmetric protocol for single-and two-qubit states. We then devise a universal protocol for arbitrary n_(1)-and n_(2)-qubit states via a(2n_(1)+2n_(2)+1)-qubit entangled state, where n_(1)≤n_(2).The receiver only needs to perform the single-qubit recovery operation, which is derived by a general expression. Moreover, a(2n_(1)+1)-bit classical communication cost can be saved within the controller’s broadcast channel by the use of network coding technology.

关键词： bidirectional controlled teleportation network coding projective measurement recovery operation

来源：评论

学校读者我要写书评

暂无评论

Fully connected-based nonnegative matrix factorization neural network 3

Fully connected-based nonnegative matrix factorization neura...

引用

3rd International Conference on Advanced algorithms and Signal Image Processing, AASIP 2023

作者： Liu, Shengnan Wei, Xiaoge Yang, Lijun School of Mathematics and Statistics Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan University Kaifeng475004 China Center for Applied Mathematics of Henan Province Henan University Zhengzhou450046 China

ISBN: (纸本)9781510668522

Nonnegative matrix factorization (NMF) is a powerful method for feature extraction, offering explanatory and dimensionality reduction. Alternatively, combining NMF with a neural network requires iterative optimization of an objective function, followed by constructing a specialized neural network based on the derived formula. The interpretability and universality of this approach are limited. To address these issues, this paper introduces a novel model called FCNMFN, which leverages a fully connected neural network to implement NMF. In this model, each layer of the fully connected neural network corresponds to the transpose of the base matrix, the coefficient matrix, and the sample matrix of NMF. This design ensures strong interpretability while achieving nonnegative matrix factorization. To demonstrate the effectiveness of the proposed model, we apply it to emotion recognition using the DEAP dataset. Experimental results confirm its efficacy and showcase its potential in accurately identifying and analyzing emotions. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

Supervised nonnegative matrix factorization model with fused correntropy for tumor recognition

Supervised nonnegative matrix factorization model with fused...

引用

2023 International Conference on Image, Signal Processing, and Pattern Recognition, ISPP 2023

作者： Wei, Xiaoge Liu, Shengnan Yang, Lijun Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms School of Mathematics and Statistics Henan University Kaifeng475004 China Center for Applied Mathematics of Henan Province Henan University Zhengzhou475004 China

ISBN: (纸本)9781510666351

Nonnegative matrix factorization (NMF) is a feature learning method that can achieve nonlinear dimensional approximate reduction with strong interpretation, and it is widely used in the field of tumor recognition. The objective function of the traditional NMF model is based on the Euclidean distance metric, and the performance of the model is easily affected by the noise. Moreover, traditional NMF is an unsupervised feature learning method that does not use the label information of the data. However, it would cause a waste of information without using label information and cannot learn the discriminative features in the data. Therefore, the supervised nonnegative matrix factorization model with fused correntropy (FCSNMF) is proposed in this paper. The FCSNMF model alleviates the effect of noise in the experimental data by fusing the Euclidean distance metric and the maximum correntropy metric. In addition, the label consistency regularization term is skillfully chosen to utilize the label information of the data to obtain discriminative features. The effectiveness of the FCSNMF model is verified by applying it to a gene expression profile dataset for tumor recognition. © 2023 SPIE.

关键词： Tumors

来源：评论

学校读者我要写书评

暂无评论

Complementary correlation and fusion of multi-modal data features for NSCLC intelligent recognition 6

Complementary correlation and fusion of multi-modal data fea...

引用

6th International Conference on Biological Information and Biomedical engineering, BIBE 2022

作者： Ma, Xiangwen Bai, Xinyu Yang, Xiaohui School of Mathematics and Statistics Henan University Kaifeng475000 China School of Mathematics and Statistics Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms Henan University Kaifeng475000 China

ISBN: (纸本)9783800759651

Recent studies have showed that the integrative analysis of CT images and genomic data can significantly improve the prediction performance of tumor diagnosis and prognosis. However, tumor classification based on the complementary correlation between CT images and genomic data has rarely been studied. In addition, when using multi-modal data features for tumor classification, most of the existing methods concatenate the multi-modal features directly, which ignore the importance of different modal features. Based on the above considerations, we construct an integrative imaging genomics classification framework for non-small cell lung cancer (NSCLC). Specifically, an adaptive weighted feature fusion method based on category contribution rate is proposed to distinguish the importance of different modal features adaptively in the classification process. Candidate’s pathogenic genes selected by correlation analysis have the potential to be suitable markers in distinguishing different subtypes of NSCLC. Our experimental results verify that our method has a better classification performance for multi-modal features than other related methods. © VDE VERLAG GMBH.

关键词： Genome

来源：评论

学校读者我要写书评

暂无评论

Unlabeled data driven cost-sensitive inverse projection sparse representation-based classification with 1/2 regularization

引用

Science China(Information Sciences) 2022年第8期65卷 39-56页

作者： Xiaohui YANG Zheng WANG Jian SUN Zongben XU Henan Engineering Research Center for Artificial Intelligence Theory and Algorithms School of Mathematics and StatisticsHenan University National Engineering Laboratory for Big Data Analytics School of Mathematics and StatisticsXi'an Jiaotong University

Sparse representation-based classification(SRC) has been widely used because it just relies on simple linear regression ideas to do classification, and it does not need learning. However, the performance of SRC is limited by needing sufficient labeled samples per class and the sensitivity to class imbalance. For tackling these problems, an improved SRC model is constructed in this paper. For alleviating the problem of insufficient labeled samples, an unlabeled data-driven inverse projection sparse representation-based classification model is constructed to achieve effective and stable representation and recognition results. The L1/2 and S1/2 regularizations are introduced to capture the sparsity of 1-D and 2-D, and to make the model have good statistical properties. Moreover, the cost-sensitive strategy is integrated into the model's classification criteria to improve the imbalance of class distribution adaptively, especially for multiclass imbalanced data.A solver utilizing the mixed Gauss-Seidel and Jacobian ADMM algorithm is developed to obtain the approximate solution. Experiments on common public test databases show that the proposed model achieves competitive results compared with the latest published results and some deep-learning algorithms.

关键词： unlabeled-data driven 1/2 regularization cost-sensitive inverse projection sparse representation-based classification

来源：评论

学校读者我要写书评

暂无评论

EEG Emotion Recognition Based on Graph Signal and Stable Learning 2

EEG Emotion Recognition Based on Graph Signal and Stable Lea...

引用

2nd International Conference on Computational Modeling, Simulation and Data Analysis, CMSDA 2022

作者： Ouyang, Rujie Liu, Shengnan Yang, Lijun School of Mathematics and Statistics Henan University Kaifeng China School of Mathematics and Statistics Henan Engineering Research Center of Artificial Intelligence Theory and Algorithm Henan University Kaifeng China

ISBN: (纸本)9798350345759

In order to extract effective information from Electroencephalography (EEG) data, this paper proposes an EEG emotion recognition method based on graph signal and stable learning. To simulate the integration of information by the brain, the EEG data is endowed with spatial structure, that is, it is modeled as a graph signal, and the graph Fourier transform is used to transform the initial graph signal into a graph signal in the frequency domain, and then the graph convolution neural network is used to automatically extract the more discriminating features of the data with graph structure for emotion classification. In addition, in order to extract the causal features in EEG channels, we used the weight learning of decorrelation to remove confounding factors and exclude correlation features. The proposed method is tested on two public datasets for binary valence and arousal, and the classification accuracy is better than other similar methods. © 2022 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

A Level Set Representation Method for N-Dimensional Convex Shape and Applications

引用

Communications in Mathematical research 2021年第2期37卷 180-208页

作者： Lingfeng Li Shousheng Luo Xue-Cheng Tai Jiang Yang Department of Mathematics of Hong Kong Baptist University Hong KongSARChina Department of Mathematics of Southern University of Science and Technology ShenzhenChina School of Mathematics and Statistics Henan Engineering Research Center for Artificial Intelligence Theory and AlgorithmsHenan UniversityKaifengChina

In this work,we present a new method for convex shape representation,which is regardless of the dimension of the concerned objects,using level-set *** the best of our knowledge,the proposed prior is the first one which can work for high dimensional *** prior is very useful for object completion in computer *** is a very challenging task to represent high dimensional convex *** this paper,we first prove that the convexity of the considered object is equivalent to the convexity of the associated signed distance ***,the second order condition of convex functions is used to characterize the shape convexity *** apply this new method to two applications:object segmentation with convexity prior and convex hull problem(especially with outliers).For both applications,the involved problems can be written as a general optimization problem with three *** algorithm based on the alternating direction method of multipliers is presented for the optimization *** experiments are conducted to verify the effectiveness of the proposed representation method and algorithm.

关键词： Convex shape prior level-set method image segmentation convex hull ADMM

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：