We present a comprehensive study on the use of autoencoders for modelling text data, in which (differently from previous studies) we focus our attention on the various issues. We explore the suitability of two differe...
详细信息
We present a comprehensive study on the use of autoencoders for modelling text data, in which (differently from previous studies) we focus our attention on the various issues. We explore the suitability of two different models binary deep autencoders (bDA) and replicated-softmax deep autencoders (rsDA) for constructing deep autoencoders for text data at the sentence level. We propose and evaluate two novel metrics for better assessing the text-reconstruction capabilities of autoencoders. We propose an automatic method to find the critical bottleneck dimensionality for text representations (below which structural information is lost);and finally we conduct a comparative evaluation across different languages, exploring the regions of critical bottleneck dimensionality and its relationship to language perplexity. (C) 2015 Elsevier B.V. All rights reserved.
As an important category of deep models, deep generative model has attracted more and more attention with the proposal of deep Belief Networks (DBNs) and the fast greedy training algorithm based on restricted Boltzman...
详细信息
As an important category of deep models, deep generative model has attracted more and more attention with the proposal of deep Belief Networks (DBNs) and the fast greedy training algorithm based on restricted Boltzmann machines (RBMs). In the past few years, many different deep generative models are proposed and used in the area of Artificial Intelligence. In this paper, three important deep generative models including DBNs, deep autoencoder, and deep Boltzmann machine are reviewed. In addition, some successful applications of deep generative models in image processing, speech recognition and information retrieval are also introduced and analysed.
The rapid advance of multimedia devices, including sensors, cameras and mobile phones, has given rise to the prevalence of Internet of Multimedia Things (IoMT), generating huge volumes of application-oriented multimed...
详细信息
The rapid advance of multimedia devices, including sensors, cameras and mobile phones, has given rise to the prevalence of Internet of Multimedia Things (IoMT), generating huge volumes of application-oriented multimedia data. At the same time, network security issues in the multimedia big data environment also increases. Network intrusion detection (NID) system demonstrates its power in preventing cyber-attacks against multimedia platforms. However, the existing NID methods which are based on machine learning or deep learning classifiers may fail when there is a lack of abnormal traffic samples for training in the real-world scenario. We propose a novel approach for intrusion detection based on deep autoencoder and Differential comparison named AED, which only requires the normal traffic samples in the training phase. We conduct extensive experiments on two real-world datasets to evaluate the effectiveness of the proposed AED. The experimental results show that AED can outperform the baseline methods of three categories in terms of accuracy, precision, recall and F1-score.
暂无评论