Filters








3,002 Hits in 3.9 sec

Relational autoencoder for feature extraction

Qinxue Meng, Daniel Catchpoole, David Skillicom, Paul J. Kennedy
2017 2017 International Joint Conference on Neural Networks (IJCNN)  
Autoencoder as a neural network based feature extraction method achieves great success in generating abstract features of high dimensional data.  ...  We also extend it to work with other major autoencoder models including Sparse Autoencoder, Denoising Autoencoder and Variational Autoencoder.  ...  Classification is done by softmax regression based on the extracted features from autoencoder models.  ... 
doi:10.1109/ijcnn.2017.7965877 dblp:conf/ijcnn/MengCSK17 fatcat:xidyrkj36zfwro6vq2afpolinu

Representation Learning with Autoencoders for Electronic Health Records: A Comparative Study [article]

Najibesadat Sadati, Milad Zafar Nezhad, Ratna Babu Chinnam, Dongxiao Zhu
2019 arXiv   pre-print
Our method uses different deep architectures (stacked sparse autoencoders, deep belief network, adversarial autoencoders and variational autoencoders) for feature representation in higher-level abstraction  ...  In this paper, we propose a predictive modeling approach based on deep learning based feature representations and word embedding techniques.  ...  In the other study [10] , a healthcare recommender system developed based on variational autoencoders and collaborative filtering.  ... 
arXiv:1908.09174v2 fatcat:67w6e435abc7xgfzobid4rweza

Representation Learning with Autoencoders for Electronic Health Records: A Comparative Study [article]

Najibesadat Sadati, Milad Zafar Nezhad, Ratna Babu Chinnam, Dongxiao Zhu
2019 arXiv   pre-print
Our method uses different deep architectures (stacked sparse autoencoders, deep belief network, adversarial autoencoders and variational autoencoders) for feature representation in higher-level abstraction  ...  In this paper, we propose a predictive modeling approach based on deep learning based feature representations and word embedding techniques.  ...  In the other study [10] , a healthcare recommender system developed based on variational autoencoders and collaborative filtering.  ... 
arXiv:1801.02961v2 fatcat:w4rqvzuvcza37hfpyqt46hazvq

Instance-Wise Denoising Autoencoder for High Dimensional Data

Lin Chen, Wan-Yu Deng
2016 Mathematical Problems in Engineering  
IDA works ahead based on the following corruption rule: if an instance vector of nonzero feature is selected, it is forced to become a zero vector.  ...  cooccurrence relation instead of the feature-wise one.  ...  Two tasks were considered: text retrieval based on an image query and image retrieval based on a query text. In the first case, each image is used as a query and produces ranking of all texts.  ... 
doi:10.1155/2016/4365372 fatcat:o227xs5jebbwbdcs3iu3envzxm

A Brief Didactic Theoretical Review on Convolutional Neural Networks, Deep Belief Networks and Stacked Auto-Encoders

Rômulo Fernandes da Costa, Sarasuaty Megume Hayashi Yelisetty, Johnny Cardoso Marques, Paulo Marcelo Tasinaffo
2019 International Journal of Engineering and Technical Research (IJETR)  
The paper focuses on the most common networks structures: Convolutional Neural Network (CNN), Deep Belief Network (DBN) and Stacked Auto-encoders (SA).  ...  The paper concludes with some considerations on the state-of-art work and on the possible future applications enabled by deep neural networks.  ...  The training in this stage is now supervised, as the labeled datais associated with the features extracted by the stack [21] . IV. STACKED AUTOENCODERS A.  ... 
doi:10.31873/ijetr.9.12.35 fatcat:4vvflbinqbh47o3x2xqqbagwae

EnsVAE: Ensemble Variational Autoencoders for Recommendations

Ahlem Drif, Houssem Eddine Zerrad, Hocine Cherifi
2020 IEEE Access  
It is based on two innovative recommender systems: 1-) a new "GloVe content-based filtering recommender" (GloVe-CBF) that exploits the strengths of embedding-based representations and stacking ensemble  ...  learning techniques to extract features from the item-based side information. 2-) a variant of neural collaborative filtering recommender, named "Gate Recurrent Unit-based Matrix Factorization recommender  ...  [12] designed a web-based movie recommender that makes suggestions based on text categorization of movie synopses.  ... 
doi:10.1109/access.2020.3030693 fatcat:wijvecqbczecrm52jlffv6q5ey

Smartphone Motion Sensor-Based Complex Human Activity Identification Using Deep Stacked Autoencoder Algorithm for Enhanced Smart Healthcare System

Uzoma Rita Alo, Henry Friday Nweke, Ying Wah Teh, Ghulam Murtaza
2020 Sensors  
Second, we propose a deep stacked autoencoder based deep learning algorithm to automatically extract compact feature representation from the motion sensor data.  ...  This paper aims to propose a deep stacked autoencoder algorithm, and orientation invariant features, for complex human activity identification. The proposed approach is made up of various stages.  ...  The deep stacked autoencoder enables extraction of more discriminant features by transforming the smartphone sensor data into hidden features in order to reduce the error rate.  ... 
doi:10.3390/s20216300 pmid:33167424 fatcat:d642qt4dwfa2pemwv4ayzk3jsa

An Extended Beta-Elliptic Model and Fuzzy Elementary Perceptual Codes for Online Multilingual Writer Identification using Deep Neural Network [article]

Thameur Dhieb, Sourour Njah, Houcine Boubaker, Wael Ouarda, Mounir Ben Ayed, Adel M. Alimi
2018 arXiv   pre-print
Then, from each stroke, we extract a set of static and dynamic features from new proposed model that we called Extended Beta-Elliptic model and from the Fuzzy Elementary Perceptual Codes.  ...  Experimental results reveal that the proposed system achieves interesting results as compared to those of the existing writer identification systems on Latin and Arabic scripts.  ...  [33] propose an approach of online text independent writer identification following four steps: features extraction using Higher Order United Moment Invariant, features ranking based on Grey Relational  ... 
arXiv:1804.05661v4 fatcat:stn222k2bvb3piozn75swir2ku

Learning Grounded Meaning Representations with Autoencoders

Carina Silberer, Mirella Lapata
2014 Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
We introduce a new model which uses stacked autoencoders to learn higher-level embeddings from textual and visual input.  ...  The two modalities are encoded as vectors of attributes and are obtained automatically from text and images, respectively.  ...  The use of stacked autoencoders to extract a shared lexical meaning representation is new to our knowledge, although, as we explain below related to a large body of work on deep learning.  ... 
doi:10.3115/v1/p14-1068 dblp:conf/acl/SilbererL14 fatcat:krrgbnx7zbcsrks6vn73432jia

The Concept of System for Automated Scientific Literature Reviews Generation [chapter]

Anton Teslyuk
2020 Lecture Notes in Computer Science  
Key elements of the system include transformer-based BERT encoder, deep LSTM decoder and a loss function which combines autoencoder loss and forces generated summaries to be in the input text domain.  ...  State of the art methods techniques based on generative adversarial networks (GANs), variational auto-encoders (VAE) and autoregressive models allow to generate images, videos, voice, texts which are very  ...  The system is based on abstractive text summarization methods: autoencoder which combines transformer BERT encoder with LSTM decoder and additional loss factor which shapes latent space to be suitable  ... 
doi:10.1007/978-3-030-50420-5_32 fatcat:zcw6jsho5jfarc77mhhsxq6mku

Learning Representations of Affect from Speech [article]

Sayan Ghosh, Eugene Laksana, Louis-Philippe Morency, Stefan Scherer
2016 arXiv   pre-print
We experiment with different input speech features (such as FFT and log-mel spectrograms with temporal context windows), and different autoencoder architectures (such as stacked and deep autoencoders).  ...  We also learn utterance specific representations by a combination of denoising autoencoders and BLSTM based recurrent autoencoders.  ...  The features extracted from the autoencoder are used for emotion classification on the IEMOCAP dataset.  ... 
arXiv:1511.04747v6 fatcat:yb76ykxyfjaqfkw43tvol4px2a

Recognition using Cyber bullying in view of Semantic-Enhanced Minimized Auto-Encoder

Jung Hyun Kim
2016 Asia-pacific Journal of Convergent Research Interchange  
Machine learning strategies make programmed identification of harassing messages in online networking conceivable, and this could develop a solid and safe web-based social networking environment.  ...  In this significant research zone, one basic issue is strong and discriminative numerical representation learning of instant messages.  ...  In this paper, we explore one profound learning strategy named stacked denoising autoencoder (SDA).  ... 
doi:10.21742/apjcri.2016.12.02 fatcat:e6hqwxrq4zcjfeopofrnhsew7m

Multimodal deep learning approach for joint EEG-EMG data compression and classification [article]

Ahmed Ben Said and Amr Mohamed and Tarek Elfouly and Khaled Harras and Z. Jane Wang
2017 arXiv   pre-print
Specifically, we build our system based on the deep autoencoder architecture which is designed not only to extract discriminant features in the multimodal data representation but also to reconstruct the  ...  Since autoencoder can be seen as a compression approach, we extend it to handle multimodal data at the encoder layer, reconstructed and retrieved at the decoder layer.  ...  [16] proposed a multimodal autoencoder [17] approach for video classification based on audio, image and text data where the intra-modality semantic for each data is separately learning by a stacked  ... 
arXiv:1703.08970v1 fatcat:sxfjktupxbdy5cj5azzwfoapxa

Multi-modal Sentiment Classification with Independent and Interactive Knowledge via Semi-supervised Learning

Dong Zhang, Shoushan Li, Qiaoming Zhu, Guodong Zhou
2020 IEEE Access  
The key idea is to leverage the semi-supervised variational autoencoders to mine more information from unlabeled data for multi-modal sentiment analysis.  ...  Multi-modal sentiment analysis extends conventional text-based definition of sentiment analysis to a multi-modal setup where multiple relevant modalities are leveraged to perform sentiment analysis.  ...  LOW-LEVEL FEATURES EXTRACTION First, we extract the low-level handcrafted features to be able to identify and interpret the factors that have the most impact on sentiment.  ... 
doi:10.1109/access.2020.2969205 fatcat:bkcydvpc7bgs3l6n4usfz7lvcy

Deep Learning for Computer Vision: A Brief Review

Athanasios Voulodimos, Nikolaos Doulamis, Anastasios Doulamis, Eftychios Protopapadakis
2018 Computational Intelligence and Neuroscience  
Denoising Autoencoders.  ...  Over the last years deep learning methods have been shown to outperform previous state-of-the-art machine learning techniques in several fields, with computer vision being one of the most prominent cases  ...  A variety of face recognition systems based on the extraction of handcrafted features have been proposed [76] [77] [78] [79] ; in such cases, a feature extractor extracts features from an aligned face  ... 
doi:10.1155/2018/7068349 pmid:29487619 pmcid:PMC5816885 fatcat:yeawpj32onfutegmkqpx4p6tsa
« Previous Showing results 1 — 15 out of 3,002 results