Commit 305a1cfa by Loïc Barrault

### update

parent f1439bb4
 ... @@ -5,7 +5,7 @@ ... @@ -5,7 +5,7 @@ \vfill \vfill \centering \centering \Huge{\edinred{[Text processing]\\Deep Learning}} \Huge{\edinred{[Text processing]\\Deep Learning for Text Processing}} \end{frame} \end{frame} ... @@ -539,36 +539,77 @@ Get a probability distribution by normalization \ra\ softmax: $p(\vc = j | \the ... @@ -539,36 +539,77 @@ Get a probability distribution by normalization \ra\ softmax:$p(\vc = j | \the \end{frame} \end{frame} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{frame} \frametitle{Deep Learning for Sentiment Analysis} \begin{block}{Principle} \myemph{Project} or represent the \textbf{text} into a \myemph{continuous space} and train an estimator operating into this space to compute the probability of the sentiment. \end{block} \begin{center} \includegraphics[height=0.6\textheight]{sa_nn} \end{center} \end{frame} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{frame} \begin{frame} \frametitle{TITLE} \frametitle{Deep Learning for NER} \begin{itemize} \item Li, J., Sun, A., Han, J., \& Li, C. (2018).\\ A Survey on Deep Learning for Named Entity Recognition. \url{http://arxiv.org/abs/1812.09449} \cite{Li2018} \end{itemize} \begin{center} \includegraphics[height=0.6\textheight]{dl_ner} \end{center} \end{frame} \end{frame} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{frame} \begin{frame} \frametitle{Deep Learning for Sentiment Analysis} \frametitle{What I couldn't talk about today} \begin{block}{Principle} \begin{itemize} \myemph{Project} or represent the \textbf{text} into a \myemph{continuous space} and train an estimator operating into this space to compute the probability of the sentiment. \item Special recurrent cells addressing the problem of \textbf{vanishing gradient} \end{block} \begin{itemize} \item LSTM: Long Short-Term Memory \cite{Hochreiter1997} \item GRU: Gated Recurrent Unit \cite{Cho2014} \end{itemize} \vfill \item Transformer models \ra\ Vaswani et al. 2017, Attention is All you Need \cite{Vaswani2017} \item[\Ra] All subsequent BERT models, see e.g. \url{https://nlp.stanford.edu/seminar/details/jdevlin.pdf} \end{itemize} \begin{center} \includegraphics[height=0.6\textheight]{sa_nn} \end{center} \end{frame} \end{frame} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{frame} \begin{frame} \frametitle{Text Processing: Deep Learning: Resources} \frametitle{Text Processing: Deep Learning: Resources} Deep Learning book: \url{https://www.deeplearningbook.org/} \begin{itemize} \item Deep Learning book: \url{https://www.deeplearningbook.org/} \cite{Goodfellow-et-al-2016} \cite{Goodfellow-et-al-2016} \vfill \item Set of notebooks for Sentiment Analysis in Pytorch\\ \url{https://github.com/bentrevett/pytorch-sentiment-analysis} \vfill \item Enjoy the lectures COM4513/6513 Natural Language Processing - Spring 19-20\\ \end{itemize} \end{frame} \end{frame} \ No newline at end of file