检索结果-内蒙古大学图书馆

Energy-efficient VM scheduling based on deep reinforcement learning

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 2021年 125卷 616-628页

作者： Wang, Bin Liu, Fagui Lin, Weiwei South China Univ Technol Sch Comp Sci & Engn Guangzhou 510006 Peoples R China

Achieving data center resource optimization and QoS guarantee driven by high energy efficiency has become a research hotspot. However, QoS information directly sampled from the cloud environment will inevitably be affected by a small amount of structured noise. This paper proposes a deep reinforcement learning model based on QoS feature learning to optimize data center resource scheduling. In the deep learning stage, we propose a QoS feature learning method based on improved stacked denoising autoencoders to extract more robust QoS characteristic information. In the reinforcement learning stage, we propose a multi-power machines (PMs) collaborative resource scheduling algorithm based on reinforcement learning. Extensive experiments show that compared with other excellent resource scheduling strategies, our method can effectively reduce the energy consumption of cloud data centers while maintaining the lowest service level agreement (SLA) violation rate. A good balance is achieved between energy-saving and QoS optimization. (C) 2021 Elsevier B.V. All rights reserved.

关键词： Energy efficiency QoS guarantee denoising autoencoder q-learning Feature learning

来源：评论

学校读者我要写书评

暂无评论

Learning social representations with deep autoencoder for recommender system

引用

WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS 2020年第4期23卷 2259-2279页

作者： Pan, Yiteng He, Fazhi Yu, Haiping Wuhan Univ Sch Comp Sci Wuhan Peoples R China

With the development of online social media, it attracts increasingly attentions to utilize social information for recommender systems. Based on the intuition that users are influenced by their social friends, these methods are capable of addressing the data sparse problem and improving the performance of recommender systems. However, these methods model the influences between each pair of users independently and ignore the interactions among these social influences, i.e., high-level signal of social information. In this paper, we propose a deep autoencoder model to learn social representations for recommender system. This approach aims to learn low- and high- level features from social information based on muti-layers neural networks and matrix factorization technique. Especially, we develop an improved deep autoencoder model, namedSparse Stacked denoising autoencoder(SSDAE), to address the data sparse and imbalance problems for social networks. Moreover, we incorporate these deep representations and matrix factorization model into a uniform framework for recommender system. Our experiments in Epinions and Ciao datasets demonstrate that our method can significantly improve the performance of recommender system, especially for sparse users.

关键词： Social recommendation Deep learning denoising autoencoder Matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Modeling for Top-N Recommendation With Interests Exploring

引用

IEEE ACCESS 2018年 6卷 51440-51455页

作者： Zhou, Wang Li, Jianping Zhang, Malu Wang, Yazhou Shah, Fadia Univ Elect Sci & Technol China Sch Comp Sci & Engn Chengdu 610054 Sichuan Peoples R China Univ Elect Sci & Technol China Sch Optoelect Informat Chengdu 610054 Sichuan Peoples R China

Recommender systems (RS) currently play a crucial role in information filtering and retrieval, and have been ubiquitously applied in many domains, although suffering from such data sparsity and cold start problems. There are plenty of studies that try to make efforts to improve the performance of RS through different aspects, such as traditional matrix factorization technique and deep learning methods in recent years, however, it's still a challenging issue under research. In this paper, motivated by this, a two-stage deep learning-based model for top-N recommendation with interests exploring (DLMR) is proposed: 1) DLMR explores latent interests for each user, captures factors from reviews and contextual information via convolutional neural network, and performs convolutional matrix factorization to generate the candidates list;2) In order to enhance the recommendation performance, DLMR further conducts candidates ranking through a three-layer denoising autoencoder, with taking account of heterogeneous side information. The DLMR provides a flexible scheme to leverage the available resources for recommendation, which is able to explore user's latent interests, capture the intricate interactions between users and items, and provide accurate and personalized recommendations. Experimental analysis over real world data sets demonstrates that DLMR could provide high performance top-N recommendation in sparse settings and outperform state-of-the-art recommender approaches significantly.

关键词： Sparse latent Dirichlet allocation interest exploring convolutional neural network denoising autoencoder top-N recommendation

来源：评论

学校读者我要写书评

暂无评论

Prediction of combustion state through a semi-supervised learning model and flame imaging

引用

FUEL 2021年 289卷 119745-119745页

作者： Han, Zhezhe Li, Jian Zhang, Biao Hossain, Md Moinul Xu, Chuanlong Southeast Univ Sch Energy & Environm Key Lab Energy Thermal Convers & Control Minist Educ Nanjing 210096 Peoples R China Univ Kent Sch Engn & Digital Arts Canterbury CT2 7NT Kent England

Accurate prediction of combustion state is crucial for an in-depth understanding of furnace performance and optimize operation conditions. Traditional data-driven approaches such as artificial neural networks and support vector machine incorporate distinct features which require prior knowledge for feature extraction and suffers poor generalization for unseen combustion states. Therefore, it is necessary to develop an advanced and accurate prediction model to resolve these limitations. This study presents a novel semi-supervised learning model integrating denoising autoencoder (DAE), generative adversarial network (GAN) and Gaussian process classifier (GPC). The DAE network is established to extract representative features of flame images and the network trained through the adversarial learning mechanism of the GAN. Structural similarity (SSIM) metric is introduced as a novel loss function to improve the feature learning ability of the DAE network. The extracted features are then fed into the GPC to predict the seen and unseen combustion states. The effectiveness of the proposed semi supervised learning model, i.e., DAE-GAN-GPC was evaluated through 4.2 MW heavy oil-fired boiler furnace flame images captured under different combustion states. The averaged prediction accuracy of 99.83% was achieved for the seen combustion states. The new states (unseen) were predicted accurately through the proposed model by fine-tuning of GPC without retraining the DAE-GAN and averaged prediction accuracy of 98.36% was achieved for the unseen states. A comparative study was also carried out with other deep neural networks and classifiers. Results suggested that the proposed model provides better prediction accuracy and robustness capability compared to other traditional prediction models.

关键词： Combustion state prediction Novel loss function denoising autoencoder Generative adversarial network Gaussian process classifier

来源：评论

学校读者我要写书评

暂无评论

Combination of bottleneck feature extraction and dereverberation for distant-talking speech recognition

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2016年第9期75卷 5093-5108页

作者： Ren, Bo Wang, Longbiao Lu, Liang Ueda, Yuma Kai, Atsuhiko Nagaoka Univ Technol 1603-1 Kamitomioka Nagaoka Niigata 9402188 Japan Univ Edinburgh Ctr Speech Technol Res Edinburgh Midlothian Scotland Shizuoka Univ Naka Ku 3-5-1 Johoku Hamamatsu Shizuoka 4328561 Japan

The performance of speech recognition in distant-talking environments is severely degraded by the reverberation that can occur in enclosed spaces (e.g., meeting rooms). To mitigate this degradation, dereverberation techniques such as network structure-based denoising autoencoders and multi-step linear prediction are used to improve the recognition accuracy of reverberant speech. Regardless of the reverberant conditions, a novel discriminative bottleneck feature extraction approach has been demonstrated to be effective for speech recognition under a range of conditions. As bottleneck feature extraction is not primarily designed for dereverberation, we are interested in whether it can compensate for other carefully designed dereverberation approaches. In this paper, we propose three schemes covering both front-end processing (cascaded combination and parallel combination) and back-end processing (system combination). Each of these schemes integrates bottleneck feature extraction with dereverberation. The effectiveness of these schemes is evaluated via a series of experiments using the REVERB challenge dataset.

关键词： Distant-talking speech recognition denoising autoencoder Bottleneck feature Dereverberation

来源：评论

学校读者我要写书评

暂无评论

An End-to-End Deep Learning Framework for Real-Time denoising of Heart Sounds for Cardiac Disease Detection in Unseen Noise

引用

IEEE ACCESS 2023年 11卷 87887-87901页

作者： Ali, Shams Nafisa Shuvo, Samiul Based Al-Manzo, Muhammad Ishtiaque Sayeed Hasan, Anwarul Hasan, Taufiq Bangladesh Univ Engn & Technol BUET Dept Biomed Engn mHlth Lab Dhaka 1205 Bangladesh Natl Heart Fdn Hosp & Res Inst Dept Cardiac Surg Dhaka 1216 Bangladesh Qatar Univ Dept Mech & Ind Engn Doha Qatar Qatar Univ Biomed Res Ctr Doha Qatar Johns Hopkins Univ Ctr Bioengn Innovat & Design Dept Biomed Engn Baltimore MD 21218 USA

The heart sound signals captured via a digital stethoscope are often distorted by environmental and physiological noise, altering their salient and critical properties. The problem is exacerbated in crowded low-resource hospital settings with high noise levels which degrades the diagnostic performance. In this study, we present a novel deep encoder-decoder-based denoising architecture (LU-Net) to suppress ambient and internal lung sound noises. Training is done using a large benchmark PCG dataset mixed with physiological noise, i.e., breathing sounds. Two different noisy datasets were prepared for experimental evaluation by mixing unseen lung sounds and hospital ambient noises with the clean heart sound recordings. We also used the inherently noisy portion of the PASCAL heart sound dataset for evaluation. The proposed framework showed effective suppression of background noises in both unseen real-world data and synthetically generated noisy heart sound recordings, improving the signal-to-noise ratio (SNR) level by 5.575 dB on an average using only 1.32 M parameters. The proposed model outperforms the current state-of-the-art U-Net model with an average SNR improvement of 5.613 dB and 5.537 dB in the presence of lung sound and unseen hospital noise, respectively. LU-Net also outperformed the state-of-the-art Fully Convolutional Network (FCN) by 1.750 dB and 1.748 dB for lung sound and unseen hospital noise conditions, respectively. In addition, the proposed denoising method model improves classification accuracy by 38.93% in the noisy portion of the PASCAL heart sound dataset. The results presented in the paper indicate that our proposed architecture demonstrated a robust denoising performance on different datasets with diverse levels and characteristics of noise. The proposed deep learning-based PCG denoising approach is a pioneering study that can significantly improve the accuracy of computer-aided auscultation systems for detecting cardiac diseases in noisy, low-

关键词： Heart sound real time denoising deep learning denoising autoencoder cardiovascular diseases cardiac disease detection

来源：评论

学校读者我要写书评

暂无评论

Deep Learning of Transferable Representation for Scalable Domain Adaptation

引用

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2016年第8期28卷 2027-2040页

作者： Long, Mingsheng Wang, Jianmin Cao, Yue Sun, Jiaguang Yu, Philip S. Tsinghua Univ Tsinghua Natl Lab Informat Sci & Techonolgy TNLis Sch Software Beijing Peoples R China Tsinghua Univ Inst Data Sci Beijing Peoples R China Univ Illinois Chicago IL 60607 USA

Domain adaptation generalizes a learning model across source domain and target domain that are sampled from different distributions. It is widely applied to cross-domain data mining for reusing labeled information and mitigating labeling consumption. Recent studies reveal that deep neural networks can learn abstract feature representation, which can reduce, but not remove, the cross-domain discrepancy. To enhance the invariance of deep representation and make it more transferable across domains, we propose a unified deep adaptation framework for jointly learning transferable representation and classifier to enable scalable domain adaptation, by taking the advantages of both deep learning and optimal two-sample matching. The framework constitutes two inter-dependent paradigms, unsupervised pre-training for effective training of deep models using deep denoising autoencoders, and supervised fine-tuning for effective exploitation of discriminative information using deep neural networks, both learned by embedding the deep representations to reproducing kernel Hilbert spaces (RKHSs) and optimally matching different domain distributions. To enable scalable learning, we develop a linear-time algorithm using unbiased estimate that scales linearly to large samples. Extensive empirical results show that the proposed framework significantly outperforms state of the art methods on diverse adaptation tasks: sentiment polarity prediction, email spam filtering, newsgroup content categorization, and visual object recognition.

关键词： Domain adaptation deep learning denoising autoencoder neural network two-sample test multiple kernel learning

来源：评论

学校读者我要写书评

暂无评论

Lifted Bregman Training of Neural Networks

引用

JOURNAL OF MACHINE LEARNING RESEARCH 2023年第1期24卷 1-51页

作者： Wang, Xiaoyu Benning, Martin Univ Cambridge Dept Appl Math & Theoret Phys Cambridge CB3 0WA England Queen Mary Univ London Sch Math Sci London E1 4NS England

We introduce a novel mathematical formulation for the training of feed-forward neural networks with (potentially non-smooth) proximal maps as activation functions. This formulation is based on Bregman distances and a key advantage is that its partial derivatives with respect to the network's parameters do not require the computation of derivatives of the network's activation functions. Instead of estimating the parameters with a combination of first-order optimisation method and back-propagation (as is the state-of-the-art), we propose the use of non-smooth first-order optimisation methods that exploit the specific structure of the novel formulation. We present several numerical results that demonstrate that these training approaches can be equally well or even better suited for the training of neural network-based classifiers and (denoising) autoencoders with sparse coding compared to more conventional training frameworks.

关键词： Lifted network training distributed optimisation Bregman distances sparse autoencoder denoising autoencoder classification compressed sensing

来源：评论

学校读者我要写书评

暂无评论

Change detection based on deep feature representation and mapping transformation for multi-spatial-resolution remote sensing images

引用

ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING 2016年第Jun.期116卷 24-41页

作者： Zhang, Puzhao Gong, Maoguo Su, Linzhi Liu, Jia Li, Zhizhou Xidian Univ Key Lab Intelligent Percept & Image Understanding Minist Educ Int Res Ctr Intelligent Percept & Computat Xian 710071 Shaanxi Provinc Peoples R China

Multi-spatial-resolution change detection is a newly proposed issue and it is of great significance in remote sensing, environmental and land use monitoring, etc. Though multi-spatial-resolution image pair are two kinds of representations of the same reality, they are often incommensurable superficially due to their different modalities and properties. In this paper, we present a novel multi-spatial resolution change detection framework, which incorporates deep-architecture-based unsupervised feature learning and mapping-based feature change analysis. Firstly, we transform multi-resolution image-pair into the same pixel-resolution through co-registration, followed by details recovery, which is designed to remedy the spatial details lost in the registration. Secondly, the denoising autoencoder is stacked to learn local and high-level representation/feature from the local neighborhood of the given pixel, in an unsupervised fashion. Thirdly, motivated by the fact that multi-resolution image-pair share the same reality in the unchanged regions, we try to explore the inner relationships between them by building a mapping neural network. And it can be used to learn a mapping function based on the most-unlikely-changed feature-pairs, which are selected from all the feature-pairs via a coarse initial change map generated in advance. The learned mapping function can bridge the different representations and highlight changes. Finally, we can build a robust and contractive change map through feature similarity analysis, and the change detection result is obtained through the segmentation of the final change map. Experiments are carried out on four real datasets, and the results confirmed the effectiveness and superiority of the proposed method. (C) 2016 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V. All rights reserved.

关键词： Change detection Spatial-resolution denoising autoencoder Stacked denoising autoencoder Deep neural networks Feature transformation

来源：评论

学校读者我要写书评

暂无评论

A novel U-Net with dense block for drum signal separation from polyphonic music signal mixture

引用

SIGNAL IMAGE AND VIDEO PROCESSING 2023年第3期17卷 627-633页

作者： George, E. Vinitha Devassia, V. P. Cochin Univ Sci & Technol Model Engn Coll Kochi Kerala India St Josephs Coll Engn & Technol Palai Kerala India

Deep neural network algorithms have shown promising results for music source signal separation. Most existing methods rely on deep networks, where billions of parameters need to be trained. In this paper, we propose a novel autoencoder framework with a reduced number of parameters to separate the drum signal component from a music signal mixture. A denoising autoencoder with a U-Net architecture and direct skip connections was employed. A dense block is included in the bottleneck of the autoencoder stage. This technique was tested on both demixing secret data (DSD) and the MUSDB database. The source-to-distortion ratio (SDR) for the proposed method was at par with that of other state-of-the-art methods, whereas the number of parameters required was quite low, making it computationally more efficient. The experiment performed using the proposed method to separate drum signal yielded an average SDR of 5.71 on DSD and 6.45 on MUSDB database while using only 0.32 million parameters.

关键词： Audio source separation Convolutional neural network Deep learning denoising autoencoder U-Net architecture

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：