检索结果-内蒙古大学图书馆

Atypical Sample Regularizer autoencoder for Cross-Domain Human Activity Recognition

INFORMATION SYSTEMS FRONTIERS 2021年第1期23卷 71-80页

作者： Prabono, Aria Ghora Yahya, Bernardo Nugroho Lee, Seok-Lyong Hankuk Univ Foreign Studies 81 Oedae Ro Yongin 17035 Gyeonggi Do South Korea

The sensor-based human activity recognition (HAR) using machine learning requires a sufficiently large amount of annotated data to realize an accurate classification model. This requirement stimulates the advancement of the transfer learning research area that minimizes the use of labeled data by transferring knowledge from the existing activity recognition domain. Existing approaches transform the data into a common subspace between domains which theoretically loses information, to begin with. Besides, they are based on the linear projection which is bound to linearity assumption and its limitations. Some recent works have already incorporated nonlinearity to find a latent representation that minimizes domain discrepancy based on an autoencoder that includes statistical distance minimization. However, such approach discovers latent representation for both domains at once, which causes sub-optimal representation because both domains compensate each other's reconstruction error during the training. We propose an autoencoder-based approach on domain adaptation for sensor-based HAR. The proposed approach learns a latent representation which minimizes the discrepancy between domains by reducing statistical distance. Instead of learning representation of both domains simultaneously, our method is a two-phase approach which first learns the representation for the domain of interest independently to ensure its optimality. Subsequently, the useful information from the existing domain is transferred. We test our approach on the publicly available sensor-based HAR datasets, using cross-domain setup. The experimental result shows that our approach significantly outperforms the existing ones.

关键词： Transfer learning autoencoder Human activity recognition Machine learning

来源：评论

学校读者我要写书评

暂无评论

An overview and empirical comparison of natural language processing (NLP) models and an introduction to and empirical application of autoencoder models in marketing

引用

JOURNAL OF THE ACADEMY OF MARKETING SCIENCE 2022年第6期50卷 1324-1350页

作者： Shankar, Venkatesh Parsana, Sohil Mays Business Sch Ctr Retailing Studies College Stn TX 77843 USA Oracle College Stn TX USA

With artificial intelligence permeating conversations and marketing interactions through digital technologies and media, machine learning models, in particular, natural language processing (NLP) models, have surged in popularity for analyzing unstructured data in marketing. Yet, we do not fully understand which NLP models are appropriate for which marketing applications and what insights can be best derived from them. We review different NLP models and their applications in marketing. We layout the advantages and disadvantages of these models and highlight the conditions under which different models are appropriate in the marketing context. We introduce the latest neural autoencoder NLP models, demonstrate these models to analyze new product announcements and news articles, and provide an empirical comparison of the different autoencoder models along with the statistical NLP models. We discuss the insights from the comparison and offer guidelines for researchers. We outline future extensions of NLP models in marketing.

关键词： Natural language processing (NLP) Topic modeling Machine learning Text analysis Text mining Unstructured data Artificial intelligence autoencoder Marketing

来源：评论

学校读者我要写书评

暂无评论

Telecom Churn Prediction Using CNN with Variational autoencoder 3rd

Telecom Churn Prediction Using CNN with Variational Autoenco...

引用

3rd International Conference on Smart Systems - Innovations in Computing (SSIC)

作者： Jain, Hemlata Khunteta, Ajay Srivastava, Sumit Poornima Univ Jaipur Comp Sci Sch Basic & Appl Sci Jaipur 303905 Rajasthan India Poornima Grp Coll Jaipur 303905 Rajasthan India Manipal Univ Jaipur Dept Informat Technol Jaipur 303007 Rajasthan India

ISBN: (纸本)9789811628771;9789811628764

Customer churn is a major problem across traditional and new age companies like telecom, digital services providers, online marketplace, payment banking and social media companies. Customer churn means customer is moving out from the company, hence impacting the top line of financial sheets. This is particularly more relevant for companies having subscription-based plans. Telecom industry which has sizable customer base and cut-throat competition is at higher risk of impact due to churn. With the customer database size running into petabyte and terabytes, it is a tedious and time-taking task to accurately predict customer churn based on some random logic. While machine learning appears to be an easier alternative, often such database contains large number of string attributes that are tough to use in machine learning algorithms. Various traditional machine learning and data mining techniques have been applied for handling such big data. All such techniques have leveraged different techniques for data engineering. In this study, we have demonstrated how a new and powerful technique convolutional neural network with variational autoencoder can help predict churn with higher accuracy. CNN automatically has the ability of good feature selection and representation of input data, and this study has come with a new aspect, i.e. integrating all string attributes in dataset (ISAD) for using all existing string variable of database helping in enhancing model performance. Experiment is done on three telecom companies Cell2Cell, Telco and Orange datasets of size 51,048, 7048 and 3333, respectively. Model outperformed on all three datasets and achieved a good accuracy level. ISAD model helped to enhance the performance by giving more feature option for predicting the customer behaviour correctly.

关键词： autoencoder Categorical features CNN Deep learning Machine learning Telecommunications Variational

来源：评论

学校读者我要写书评

暂无评论

A Convolutional autoencoder based Keyword Spotting in Historical Handwritten Devanagari Documents 5

A Convolutional Autoencoder based Keyword Spotting in Histor...

引用

5th International Conference on Inventive Computation Technologies (ICICT)

作者： Sushma, S. N. Sharada, B. Univ Mysore Comp Sci Dept Mysore Karnataka India

ISBN: (数字)9781665408370

ISBN: (纸本)9781665408387;9781665408370

There exist huge amount of valuable historical Devanagari documents archived in many national libraries that need to be preserved in digital form. Additionally, there is a growing requirement in the area of document image processing for automation and information extraction from old handwritten documents. Retrieval of relevant information from handwritten historical document images would ideally necessitate an efficient An alternate approach to accurate document transcription is to use a keyword spotting method. The keyword spotting system's primary application is to create digital libraries that facilitate quick searching and browsing of old manuscripts in order to preserve the world's cultural heritage. For word spotting in handwritten manuscripts, we propose a convolution neural network (CNN) architecture based on autoencoder representation. The proposed technique was demonstrated using historical handwritten Devanagari manuscript which is collected from Oriental Research Institute of University of Mysore, Mysore. As a result, the proposed convolution neural network (CNN) method exhibits superior accuracy with favorable results.

关键词： Keyword spotting Convolutional neural network autoencoder

来源：评论

学校读者我要写书评

暂无评论

Astral: An autoencoder-Based Model for Pedestrian Trajectory Prediction of Variable-Length

Astral: An Autoencoder-Based Model for Pedestrian Trajectory...

引用

27th International Conference on Database Systems for Advanced Applications (DASFAA)

作者： Diao, Yupeng Su, Yiteng Zeng, Ximu Chen, Xu Liu, Shuncheng Su, Han Univ Elect Sci & Technol China Sch Comp Sci & Engn Chengdu Peoples R China Univ Elect Sci & Technol China Yangtze Delta Reg Inst Quzhou Chengdu Peoples R China

ISBN: (纸本)9783031112171;9783031112164

Predicting pedestrian's future behavior in a crowd plays an important role in many fields. Such as autonomous driving, machine navigation, video surveillance, and intelligent security systems. This is very challenging because pedestrian motion can be easily influenced by surrounding pedestrians' interactions. In previous works, researchers use these interactions to make prediction more effective. However, the previous work set fixed-length input in their models. In this way, they ignore shorter pedestrian trajectories. This approach leads to insufficient feature information and inaccurately prediction in some scenarios. In this paper, we propose an autoencoder-based model for pedestrian trajectory prediction of variable length (ASTRAL). At first, we use the autoencoder to process pedestrian data with variable-length trajectories. And then, we use the optimized multi-head attention mechanism to extract the interactions between neighbors. Finally, we use LSTM to decode vectors and make predictions. In particular, we fine-tune the model to make its performance better. We test our model and the state-of-the-art methods on the public benchmark datasets. Compared with others, our model improves ADE (average displacement error) and FDE (final displacement error) by 9% and 33% respectively. Therefore, our model is better than previous works, and we can predict the future trajectory of pedestrians more effectively.

关键词： Variable length trajectory autoencoder Multi-head attention

来源：评论

学校读者我要写书评

暂无评论

FrameCorr: Adaptive, autoencoder-based Neural Compression for Video Reconstruction in Resource and Timing Constrained Network Settings 26

FrameCorr: Adaptive, Autoencoder-based Neural Compression fo...

引用

26th International Symposium on Multimedia, ISM 2024

作者： Li, John Nair, Deepak Nahrstedt, Klara Gupta, Indranil Ahmed, Shehab Sarar University of Illinois Urbana-Champaign United States

ISBN: (纸本)9798331511111

Video processing is becoming increasingly popular and cost-effective on IoT devices but faces challenges in transmitting data under varying timing constraints and network bandwidth. Existing compression methods struggle with incomplete data. We present FrameCorr, a deep learning framework that leverages prior-received video data to predict and reconstruct missing frame segments, enabling video reconstruction despite data loss. © 2024 IEEE.

关键词： autoencoder IoT Neural Compression Video Transmission

来源：评论

学校读者我要写书评

暂无评论

Computationally Efficient Hyperspectral Image Classification Using Deep autoencoder and Multistage 2 Dimensional Convolutional autoencoder Architectures

引用

Procedia Computer Science 2025年 260卷 1017-1025页

作者： Sneha R. Burnase Poonam N. Sonar Maitreya Shelare Rajiv Gandhi Institute of Technology

Hyper-spectral imaging (HSI) sensor technology enables extensive coverage with spatial, spectral, and temporal flexibility. However, the substantial volume of spectral-spatial information it provides poses significant challenges in data management and extraction. To address these challenges, various advancements have been made in hyperspectral imaging over recent decades. Key issues in hyperspectral image Classification include the large data size, insufficient training samples, particularly the scarcity of labeled samples and computational burden effective feature extraction methods and the appropriate selection of filters for classification are critical for achieving reliable results without data loss. In this paper, two methods named Deep autoencoder (DAE) and Multistage 2D- Convolutional autoencoder (2D-CAE) proposed with simplified and robust architecture which helped in reduced space and time complexity and efficient in dimentionality reduction. Analysed comparison for Support Vector Machine (SVM), K- nearest Neighbour (KNN), HybridSN, Wavelet-CNN is done with proposed methods (DAE) and Multistage 2D- Convolutional autoencoder (2D- CAE), evaluating their accuracy and computational efficiency with respect to Overall Accuracy (OA), Average Accuracy(AA), Kappa(k), FLOPs, Throughput and Test time required for prediction by performing extensive experiments on publically available dataset Indian Pines (IP), Salina (SA) and Pavia University (PU). FLOPs reduction in DAE is 99.97% and 2D-CAE is 99.95%, throughput reduction in DAE is 99.95% and 2D-CAE is 99.98% when compared individually with HybridSN and Wavelet-CNN, showing significant reduction in space complexity.

关键词： autoencoder Classification Convolution Complexity Hyper-spectral

来源：评论

学校读者我要写书评

暂无评论

Classification of recurrent major depressive disorder using a residual denoising autoencoder framework: Insights from large-scale multisite fMRI data

引用

COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2024年 247卷 108114-108114页

作者： Dai, Peishan Shi, Yun Lu, Da Zhou, Ying Luo, Jialin He, Zhuang Chen, Zailiang Zou, Beiji Tang, Hui Huang, Zhongchao Liao, Shenghui Cent South Univ Sch Comp Sci & Engn Changsha 410083 Hunan Peoples R China Cent South Univ Xiangya Hosp 2 Dept Psychiat Changsha 410083 Hunan Peoples R China Cent South Univ Xiangya Hosp 2 Natl Clin Res Ctr Mental Disorders Changsha 410083 Hunan Peoples R China Cent South Univ Sch Basic Med Sci Dept Biomed Engn Changsha 410083 Hunan Peoples R China

Background and objective: Recurrent major depressive disorder (rMDD) has a high recurrence rate, and symptoms often worsen with each episode. Classifying rMDD using functional magnetic resonance imaging (fMRI) can enhance understanding of brain activity and aid diagnosis and treatment of this disorder. Methods: We developed a Residual Denoising autoencoder (Res-DAE) framework for the classification of rMDD. The functional connectivity (FC) was extracted from fMRI data as features. The framework addresses site heterogeneity by employing the Combat method to harmonize feature distribution differences. A feature selection method based on Fisher scores was used to reduce redundant information in the features. A data augmentation strategy using a Synthetic Minority Over -sampling Technique algorithm based on Extended Frobenius Norm measure was incorporated to increase the sample size. Furthermore, a residual module was integrated into the autoencoder network to preserve important features and improve the classification accuracy. Results: We tested our framework on a large-scale, multisite fMRI dataset, which includes 189 rMDD patients and 427 healthy controls. The Res-DAE achieved an average accuracy of 75.1 % (sensitivity = 69 %, specificity = 77.8 %) in cross-validation, thereby outperforming comparison methods. In a larger dataset that also includes first-episode depression (comprising 832 MDD patients and 779 healthy controls), the accuracy reached 70 %. Conclusions: We proposed a deep learning framework that can effectively classify rMDD and 33 identify the altered FC associated with rMDD. Our study may reveal changes in brain function 34 associated with rMDD and provide assistance for the diagnosis and treatment of rMDD.

关键词： Recurrent major depressive disorder Functional magnetic resonance imaging Functional connectivity autoencoder Residual network Deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Plant Disease Detection in Precision Agriculture Using autoencoder and Transfer Learning 3

Enhanced Plant Disease Detection in Precision Agriculture Us...

引用

3rd International Conference on Artificial Intelligence, Computational Electronics and Communication System, AICECS 2024

作者： Prema Kumar, M. Gunasundari, B. Pradeepa, R. Sulthana, K.Fouzia Lavanya, Tati Philip, Jeethu Shri Vishnu Engineering College for Women (A) Dept.of ECE Bhimavaram India Prathyusha Engineering College Department of CSE Thiruvallur India Sona College of Technology Department of IT Salem India Rajalakshmi Institute of Technology Department of AI&DS Chennai India Malla Reddy Engineering College Department of CSE (Data Science) Maisammaguda India MLR Institute of Technology Department of IT Hyderabad India

ISBN: (纸本)9798350391244

Agricultural productivity faces significant threats from plant diseases, leading to substantial economic losses and worsening global food insecurity. These diseases can drastically reduce crop yields, adversely affecting farmers' livelihoods and disrupting food supply chains. In response, agrotechnology and precision farming have emerged as critical solutions, emphasizing the need for sustainable crop management through accurate and timely disease detection. Early intervention is essential to minimize disease spread and enhance understanding of crop health. This study presents a comprehensive approach to plant disease classification using a combination of autoencoders for dimensionality reduction and a fine-tuned ResNet-9 architecture. The proposed methodology leverages the autoencoder to effectively reduce the dimensionality of leaf imagery data, thereby simplifying the feature space and improving computational efficiency. The reduced features are then fed into a ResNet-9 model, which has been fine-tuned to achieve high classification accuracy for various plant diseases. Our model achieves a remarkable accuracy of 99.2%. The development of this robust tool aims to facilitate early disease detection and improve crop management, contributing significantly to the field of precision agriculture. By providing precise and timely diagnoses ofplant health, deep learning models hold the potential to revolutionize agricultural practices, thereby enhancing yield and sustainability in food production. © 2024 IEEE.

关键词： autoencoder dimensionality reduction plant disease Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Radar pre-sorting algorithm based on autoencoder and LSTM

引用

AEU-INTERNATIONAL JOURNAL OF ELECTRONICS AND COMMUNICATIONS 2024年 187卷

作者： Jiang, Yilin Shi, Shaoxiong Zhang, Fangyuan Huang, Wuqi Harbin Engn Univ Coll Informat & Commun Engn Harbin 150001 Peoples R China Harbin Univ Sci & Technol Harbin Peoples R China AVIC United Technol Ctr Electromagnet Spectrum Col Harbin Peoples R China

As the electromagnetic environment becomes increasingly complex, most current radar signal sorting methods are unsustainable. They often perform poorly when dealing with unknown radar types and low-frequency radar pulse data. This paper introduces a radar pre-sorting algorithm based on autoencoder and LSTM. The algorithm utilizes multi-dimensional information such as pulse width, carrier frequency, and time of arrival. The autoencoder network is employed to achieve automatic feature extraction and clustering, enhancing the extraction of latent features in the data. The proposed network model mainly consists of three parts: an encoding module composed of a convolutional neural network (CNN), a feature aggregation module composed of long short-term memory (LSTM), and a decoding module obtained through a convolutional autoencoder, referred to as CLDE (CNN-LSTM-Decode). The encoding module extracts features from multi-dimensional data to obtain compressed features, the feature accumulation module processes the compressed features, further extracting hidden features between pulses. Subsequently, the decoding module determines the pulse modulation type of each pulse, achieving the purpose of radar pulse signal pre-sorting. Simulation results show that this network structure effectively pre-classifies unknown radar signals and has a high recognition rate for low-frequency pulses. Additionally, CLDE exhibits high reliability and stability in environments with pulse loss.

关键词： CLDE Low-frequency autoencoder LSTM

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：