Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Anomaly detection refers to the problem of finding rare patterns in data which raise suspicions because they do not comply with an expected behavior. We can consider different kinds of applications like intrusion detection, image processing, system health monitoring and sensor networks. For example, an anomalous pattern coming from sensors on a machine could mean that the machine is ready to break. 
Most of the current studies on anomaly detection do not consider recent/past events to detect possible new incoming outliers. The use of Long Short-Term Memory (LSTM) networks is then proposed to deal with time dependent data related with anomaly detection problems.
The goal of Support Vector Data Description (SVDD) is to describe a realistic domain for the data, excluding superfluous space. The resulting boundary can then be used to detect outliers.

In this master thesis, we consider a LSTM-based prediction model for sensor readings coming from a pulp and paper manufacturing machine. Anomalies will then result from too large prediction errors. We compare the SVDD and a discrimination rule based on the assumption of normality for the errors. In the final chapter, we show that for a real world applications the Gaussian distribution for the errors cannot hold and that the need of a non-parametric data descriptions using kernels is real.

Long short term memory networks --- Anomaly detection --- Support vector data description --- LSTM --- SVDD --- Neural networks --- Vanishing gradient problem --- Auto-Encoder-Decoder --- Physique, chimie, mathématiques & sciences de la terre > Mathématiques

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The stratospheric ozone is important for the protection of the biosphere from the dangerous ultraviolet radiation of the sun, forms the temperature and dynamical structure of the stratosphere, and, therefore, has a direct influence on the general circulation and the surface climate. The tropospheric ozone can damage the biosphere, impact human health, and plays a role as a powerful greenhouse gas. That is why the understanding of the past and future evolution of the ozone in different atmospheric layers, as well as its influence on surface UV radiation doses, and human health is important. The problems of preventing further destruction of the ozone layer, the restoration of the ozone shield in the future, and air quality remain important for society. The interest in these problems was recently enhanced by the unexpected discovery of a negative ozone trend in the lower stratosphere and the appearance of a large ozone hole over the Arctic in spring 2020. This book includes papers describing several aspects of the ozone layer’s state and evolution based on the recent experimental, statistical, and modeling works. The book will be useful for readers, scientists, and students interested in environmental science.

ozone --- PM2.5 --- PM10 --- nitrogen dioxide --- respiratory disease --- decision tree model --- merra ozone data --- discontinuities in reanalysis time series --- trend analyses --- total ozone content --- cloudiness --- erythemal radiation --- trend --- chemical–climate model --- ERA-Interim reanalysis --- Northern Eurasia --- UV resources --- stratospheric ozone --- natural and anthropogenic factors --- numerical modeling --- satellite observations --- trend estimations --- tropospheric ozone --- stratospheric intrusion --- horizontal-trough --- ozone layer evolution --- modeling --- climate change --- solar forcing --- ozone precursors --- total column of ozone (TCO) --- trend estimates --- long short-term memory networks (LSTM) --- empirical wavelet transform (EWT) --- forecasting --- Mann-Kendall --- ozone exceedance --- urban site --- rural site --- human health --- ozone enhancement --- Irene --- ozone decline --- potential vorticity --- ozonesondes --- ultraviolet radiation --- forcing

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The introduction of new topologies and training procedures to deep neural networks has solicited a renewed interest in the field of neural computation. The use of deep structures has significantly improved state-of-the-art applications in many fields, such as computer vision, speech and text processing, medical applications, and IoT (Internet of Things). The probability of a successful outcome from a neural network is linked to selection of an appropriate network architecture and training algorithm. Accordingly, much of the recent research on neural networks has been devoted to the study and proposal of novel architectures, including solutions tailored to specific problems. This book gives significant contributions to the above-mentioned fields by merging theoretical aspects and relevant applications.

Information technology industries --- facial image analysis --- facial nerve paralysis --- deep convolutional neural networks --- image classification --- Chinese text classification --- long short-term memory --- convolutional neural network --- Arabic named entity recognition --- bidirectional recurrent neural network --- GRU --- LSTM --- natural language processing --- word embedding --- CNN --- object detection network --- attention mechanism --- feature fusion --- LSTM-CRF model --- elements recognition --- linguistic features --- POS syntactic rules --- action recognition --- fused features --- 3D convolution neural network --- motion map --- long short-term-memory --- tooth-marked tongue --- gradient-weighted class activation maps --- ship identification --- fully convolutional network --- embedded deep learning --- scalability --- gesture recognition --- human computer interaction --- alternative fusion neural network --- deep learning --- sentiment attention mechanism --- bidirectional gated recurrent unit --- Internet of Things --- convolutional neural networks --- graph partitioning --- distributed systems --- resource-efficient inference --- pedestrian attribute recognition --- graph convolutional network --- multi-label learning --- autoencoders --- long-short-term memory networks --- convolution neural Networks --- object recognition --- sentiment analysis --- text recognition --- IoT (Internet of Thing) systems --- medical applications

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The introduction of new topologies and training procedures to deep neural networks has solicited a renewed interest in the field of neural computation. The use of deep structures has significantly improved state-of-the-art applications in many fields, such as computer vision, speech and text processing, medical applications, and IoT (Internet of Things). The probability of a successful outcome from a neural network is linked to selection of an appropriate network architecture and training algorithm. Accordingly, much of the recent research on neural networks has been devoted to the study and proposal of novel architectures, including solutions tailored to specific problems. This book gives significant contributions to the above-mentioned fields by merging theoretical aspects and relevant applications.

Information technology industries --- facial image analysis --- facial nerve paralysis --- deep convolutional neural networks --- image classification --- Chinese text classification --- long short-term memory --- convolutional neural network --- Arabic named entity recognition --- bidirectional recurrent neural network --- GRU --- LSTM --- natural language processing --- word embedding --- CNN --- object detection network --- attention mechanism --- feature fusion --- LSTM-CRF model --- elements recognition --- linguistic features --- POS syntactic rules --- action recognition --- fused features --- 3D convolution neural network --- motion map --- long short-term-memory --- tooth-marked tongue --- gradient-weighted class activation maps --- ship identification --- fully convolutional network --- embedded deep learning --- scalability --- gesture recognition --- human computer interaction --- alternative fusion neural network --- deep learning --- sentiment attention mechanism --- bidirectional gated recurrent unit --- Internet of Things --- convolutional neural networks --- graph partitioning --- distributed systems --- resource-efficient inference --- pedestrian attribute recognition --- graph convolutional network --- multi-label learning --- autoencoders --- long-short-term memory networks --- convolution neural Networks --- object recognition --- sentiment analysis --- text recognition --- IoT (Internet of Thing) systems --- medical applications