检索结果-内蒙古大学图书馆

3rd International Conference on Electronic Information Engineering and Computer Science, EIECS 2023

作者： Li, Jiatu University of California San Diego Halicioǧlu Data Science Institute Department of Mathematics San DiegoCA United States

ISBN: (纸本)9798350316674

Accurate medical imaging is vital for precise disease diagnosis and effective treatment. However, X-ray images may be subject to varying degrees of noise due to factors such as patient health conditions requiring reduced X-ray radiation. In this paper, the Medical Image denoising and Disease Classification Network (MIDC-Net) is proposed to solve the issue of noisy chest X-rays that negatively affects thorax disease classification by performing image denoising and disease classification simultaneously. The MIDC-Net integrates a Convolutional Neural Network (CNN) for classification, and a denoising autoencoder with convolutional layers and residual connections for image denoising. It comprises three components, namely an encoder, a decoder, and a classifier, where the input of the encoder is noisy images with varying levels of Gaussian noise, and the output of the decoder and the classifier are reconstructed images and predicted diseases. The loss of the MIDC-Net is a linear combination of Mean Square Error (MSE) loss used for denoising and Binary Cross-Entropy (BCE) loss used for multi-label classification. Experimental outcomes on the NIH chest X-ray dataset illustrated that the reconstructed images resembled the original X-rays, and the network achieved a low MSE loss, but the quality of denoised images exhibited an inverse relationship with noise levels. Regarding multi-label thorax classification results, the test accuracies all exceed 95% regardless of noise levels of the input images, demonstrating efficient and accurate prediction in the scenario of noisy inputs. © 2023 IEEE.

关键词： convolutional neural network denoising autoencoder Image denoising multi-label classification

来源：评论

学校读者我要写书评

暂无评论

LPI radar signal recognition with U2-Net-based denoising 14

LPI radar signal recognition with U2-Net-based denoising

引用

14th International Conference on Information and Communication Technology Convergence, ICTC 2023

作者： Lee, Siho Nam, Haewoon Hanyang University Department of Electrical and Electronic Engineering Ansan Korea Republic of

ISBN: (纸本)9798350313277

Low Probability of Intercept (LPI) radar signals play a vital role in electronic warfare by maintaining informational superiority. Classifying these LPI radar waveforms is a key capability but remains a challenging task due to strong noise interference. Traditional signal processing techniques often show limitations in effectively removing complex noise signals. While deep learning-based modulation classification has exhibited superior performance, its effectiveness is compromised in the presence of significant noise. In this study, we propose a deep learning-based denoising method using the U2-Net for LPI radar signals, followed by modulation classification using a Convolutional Neural Network (CNN). We further compare the performance of U2-Net with other denoising models such as U-Net and denoising autoencoder. Experimental results demonstrate that the U2-Net outperforms other methods, achieving over 90% classification accuracy for signals with a signal-to-noise ratio above -14dB. © 2023 IEEE.

关键词： denoising autoencoder Low Probability of Intercept (LPI) radar time frequency analysis U-Net U2-Net

来源：评论

学校读者我要写书评

暂无评论

NrGe-DTL: a computational framework for cancer drug response prediction based on deep transfer learning from combined denoised genomic profiles and chemical structure embedding of drugs

NrGe-DTL: a computational framework for cancer drug response...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Zhang, Yuchen Lian, Linghang Yang, Xuhua Zhejiang University of Technology Coll. of Computer Science and Technology Hangzhou China

ISBN: (纸本)9798350337488

In recent years, precision medicine has been consistently studied and employed in cancer treatment. One of the main challenges in precision medicine is accurately predicting a cancer patient's response to a specific drug(s) using computational models. Due to the heterogeneity of cancer, the distribution between the drug response profiles based on cell lines and real samples is often different, which leads to out-of-data distribution problems that affect putting the existing predictive models into practical application. To overcome this limitation, a novel framework to predict cancer drug response based on transfer learning from combined genomic profiles and chemical structure embedding of drugs is proposed in this paper. This framework features a denoising auto-encoder (DAE) module designed to remove noise from genomics profiles and a graph convolutional network (GCN) employed to extract the embedding of pharmaceutical chemical structures. Through the experiments, our model outperforms recently published methods on Genomics of Drug Sensitivity in Cancer (GDSC) datasets, and gives promising results on personalized drug response prediction on TCGA pan-cancer datasets, especially for the response of two commonly used anti-cancer drugs―Gemcitabine and Cisplatin. In summary, our study designed a computational framework with the potential capability of predicting personalized cancer drug response, which has been preliminarily validated for its effectiveness. © 2023 IEEE.

关键词： chemical structure information denoising autoencoder drug response precision medicine transfer learning

来源：评论

学校读者我要写书评

暂无评论

Denoised Labels for Financial Time Series Data via Self-Supervised Learning 22

Denoised Labels for Financial Time Series Data via Self-Supe...

引用

3rd ACM International Conference on AI in Finance (ICAIF)

作者： Ma, Yanqing Ventre, Carmine Polukarov, Maria Kings Coll London London England

ISBN: (纸本)9781450393768

The introduction of electronic trading platforms effectively changed the organisation of traditional systemic trading from quote-driven markets into order-driven markets. Its convenience led to an exponentially increasing amount of financial data, which is however hard to use for the prediction of future prices, due to the low signal-to-noise ratio and the non-stationarity of financial time series. Simpler classification tasks - where the goal is to predict the directions of future price movement via supervised learning algorithms need sufficiently reliable labels to generalise well. Labelling financial data is however less well defined than in other domains: did the price go up because of noise or a signal? The existing labelling methods have limited countermeasures against the noise, as well as limited effects in improving learning algorithms. This work takes inspiration from image classification in trading [6] and the success of self-supervised learning in computer vision (e.g., [16]). We investigate the idea of applying these techniques to financial time series to reduce the noise exposure and hence generate correct labels. We look at label generation as the pretext task of a self-supervised learning approach and compare the naive (and noisy) labels, commonly used in the literature, with the labels generated by a denoising autoencoder for the same downstream classification task. Our results demonstrate that these denoised labels improve the performances of the downstream learning algorithm, for both small and large datasets, while preserving the market trends. These findings suggest that with our proposed techniques, self-supervised learning constitutes a powerful framework for generating "better" financial labels that are useful for studying the underlying patterns of the market.

关键词： Self-Supervised Learning denoising autoencoder Financial Labels Directional-Change Events

来源：评论

学校读者我要写书评

暂无评论

Prediction model of bearing fault remaining useful life based on weighted variable loss degradation characteristics

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2024年第9期35卷 096122-096122页

作者： Yu, Tianyi Li, Shunming Lu, Jiantao Nanjing Univ Aeronaut & Astronaut Coll Energy & Power Engn Nanjing 210016 Peoples R China Nantong Inst Technol Sch Automot Engn Nantong 226002 Peoples R China

In the prediction of bearing fault remaining useful life (RUL), the identification and feature extraction of early bearing faults are very important. In order to improve the accuracy of early fault RUL prediction, a bearing fault RUL prediction model based on weighted variable loss degradation characteristics is proposed. The model is composed of a stack denoising autoencoder (SDAE) module guided by variable loss, a signal-to-noise feature adaptive weighting module and a long-short term memory (LSTM) degradation characteristics extraction and regression output module. Firstly, this model improves the ability of SDAE model to extract weak fault features by ascending dimension learning and variable loss function. Then, an adaptive weighting matrix is generated according to the test signal to modulate the weight vector of SDAE. Finally, the hidden layer features of SDAE were input into LSTM model to extract the bearing state degradation features and realize the RUL prediction of bearing faults. The experimental results show that the proposed model can accurately predict the RUL of the test data in the early fault stage and the fault development stage. The proposed model can give early fault warning to the bearing state.

关键词： remaining useful life prediction early failure denoising autoencoder long-short term memory bearing failure

来源：评论

学校读者我要写书评

暂无评论

A novel health indicator by dominant invariant subspace on Grassmann manifold for state of health assessment of lithium-ion battery

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2024年 130卷

作者： Zhang, Ying Li, Yan-Fu Zhang, Ming Wang, Huan Univ Sci & Technol Beijing Sch Mech Engn Beijing 100084 Peoples R China Tsinghua Univ Dept Ind Engn Beijing 100084 Peoples R China Tsinghua Univ Inst Qual & Reliabil Beijing 100084 Peoples R China Aston Univ Coll Engn & Phys Sci Birmingham B4 7ET England

The precise estimation of the state of health (SoH) in Lithium-ion batteries (LiBs) relies heavily on a reliable health indicator (HI). Conventional indicators are often constructed by directly concatenating features from multiple sources. It overlooks significant non-linear and correlative information inherent in raw signals. To address this limitation, this paper introduces an innovative approach for SoH estimation in LiBs. Deep features extracted from signals of various sensors are obtained using denoising auto-encoders (DAEs). Then the dominant invariant subspaces (DIS) are calculated through the non-linear transformation of multi-source features on the Grassmann manifold. It can preserve essential and robust characteristics. The health indicator quantifies the geodesic distance of DIS using a projection metric. It provides a more comprehensive inclusion of nonlinear and correlation information. Consequently, this indicator offers heightened precision in discerning differences in health states. Validation of the proposed method is conducted using the NASA dataset. The result demonstrates its effectiveness on the SoH assessment and superiority to the state-of-the-art method.

关键词： State of health Lithium-ion battery Grassmann manifold Dominant invariant subspace Health indicator denoising autoencoder Multi-source information fusion

来源：评论

学校读者我要写书评

暂无评论

Noise reduction method for wind turbine gearbox vibration signals based on CVMD-DRDSAE

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2024年第11期35卷 116146-116146页

作者： Yao, Jinbao Yue, Bohao Wang, Yizhu Li, Xiang Southwest Petr Univ Sch Mechatron Engn Chengdu 610500 Peoples R China

Wind turbine gearbox fault feature extraction is difficult due to strong background noise. To address this issue, a noise reduction method combining comprehensive learning particle swarm optimization-variational mode decomposition (CLPSO-VMD) and deep residual denoising self-attention autoencoder (DRDSAE) is proposed. Firstly, the proposed CLPSO-VMD algorithm is used to decompose the noisy wind turbine gearbox vibration signals. Subsequently, the intrinsic mode functions highly correlated with the original signals are selected through the Spearman correlation coefficient and utilized for signal reconstruction, thereby filtering out high-frequency noise outside the fault frequency band in the frequency domain characterization. Secondly, the improved DRDSAE is utilized to learn the latent representations of data in the first-level denoised signal, further reducing the residual noise within the fault frequency band while retaining important signal features. Finally, the envelope spectrum highlights the weak feature of the wind turbine gearbox vibration signal. Experimental results demonstrate the effectiveness of the proposed method in denoising wind turbine gearbox vibration signals under strong noise.

关键词： wind turbine gearbox signal denoising variational mode decomposition self-attention mechanism denoising autoencoder

来源：评论

学校读者我要写书评

暂无评论

Cleaning Noisy ECG based on the Signal Quality with Single and Multiple Hidden Layer autoencoder 2

Cleaning Noisy ECG based on the Signal Quality with Single a...

引用

2nd International Conference on Intelligent Cybernetics Technology and Applications (ICICyTA) - Cybernetic Technology for Better Future of Human Life

作者： Marwan, Bagas Samann, Fars Schanze, Thomas Univ Appl Sci THM Fac Life Sci Engn LSE Inst Biomed Engn IBMT Giessen Germany Univ Duhok Dept Biomed Engn Duhok Kurdistan Regio Iraq

ISBN: (纸本)9798350399134

Bio-signal recordings, e.g., electrocardiogram (ECG), often interfere with noise from different sources, which could affect the morphology of essential features, such as P-wave and T-wave. These features are superimposed with noise and make it difficult to interpret them. Here, we target to improve the quality of the ECG features with unsupervised neural networks, such as denoising autoencoder (DAE). In this work, the training and testing performance of single hidden layer and multiple hidden layers denoising autoencoders (SHL- and MHL-DAE) is evaluated based on the signal quality and the restoration of the ECG morphology waveforms by using signal-to-noise ratio improvement and the characteristic morphology of the ECG waves. The DAE was trained and tested with the superposition of simulated ground truth ECG and recorded physical noises, such as baseline wandering noise, electrode movement noise, muscle artifacts noise, a mixture of noises, and gaussian white noise. The effects of using different activation functions, the number of neurons, the number of the dataset, and dataset pre-processing are presented in this work. In the case of jittered-segment datasets, the MHL-DAE performs better than the SHL-DAE. However, the selection of the essential activation function and set of hyperparameters, such as the number of hidden neurons, must be done carefully.

关键词： denoising autoencoder filtering activation function hidden neurons jittering physical noise

来源：评论

学校读者我要写书评

暂无评论

Data-driven Prior for Pharmaceutical Snapshot Spectral Imaging 25

Data-driven Prior for Pharmaceutical Snapshot Spectral Imagi...

引用

25th IEEE International Conference on Computational Science and Engineering (CSE)

作者： Su, Xuesan Mao, Jianxu Wang, Yaonan Chen, Yurong Zhang, Hui Hunan Univ NERC RVC Changsha Peoples R China

ISBN: (纸本)9798350396331

This paper proposes a new method for pharmaceutical hyperspectral compressive imaging and has a significant improvement for the quality of reconstruction. It's known that coded aperture snapshot spectral imager(CASSI) overcomes the limitation of hyperspectral image acquisition. However, the spatial and spectral information is coded and overlapped which make it difficult to reconstruct the original images. The reconstruction is an inverse mathematical problem which is barely solved precisely especially in complex imaging scenes such as irregular pharmaceutical product imaging. Thus, we consider the real pharmaceutical imaging demands and propose a novel image restoration method with the data-driven prior. Our method is based on the generalized alternating projection(GAP) framework and propose a novel denoising part to solve the problem of detail texture feature extraction with the dense block module employed. Our method is tested on real pharmaceutical hyperspectral data and achieve higher performance compared with state of the art methods.

关键词： computational imaging compressive sensing denoising autoencoder dense block data-driven prior

来源：评论

学校读者我要写书评

暂无评论

Design of Reconfigurable Metasurface Based on Support Vector Machine

Design of Reconfigurable Metasurface Based on Support Vector...

引用

IEEE MTT-S International Microwave Workshop Series on Advanced Materials and Processes for RF and THz Applications (IMWS-AMP)

作者： Lu, Ying Juan Gao, Shang Wang, Zheng Xing Liang, Jing Cheng Cheng, Qiang Southeast Univ State Key Lab Millimeter Waves Nanjing 210018 Peoples R China

ISBN: (纸本)9781665478342

Traditional metasurface design requires strong expert experience and large time cost. To overcome this problem, a fast design method for reconfigurable metasurface based on support vector machine is proposed, which can be divided into four stages: data collection, data preprocessing, model training, and reverse prediction. Based on the improved pixel modeling method, 10,000 groups of data concerning the reflection amplitude for active unit cells with different patterns and excitations are collected as a dataset. denoising autoencoder is used to extract the features of the data, with fitting errors below 0.0053, whose accuracy is an order of magnitude higher than other autoencoders. To reduce the cost of parameter optimization and improve the element performance, particle swarm optimization is introduced into the optimization process of the SVM. The predicted result is in good agreement with the design target, which verifies the accuracy of the rapid design model.

关键词： Support Vector Machine denoising autoencoder Reconfigurable Intelligent Surfaces Particle Swarm Optimization

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：