Added a few lines

Author: Vineet John

project-report/cs698_project_report.tex

@@ -77,7 +77,13 @@ \section{A Primer of Neural Net Models for NLP} % (fold)
         \item 
         The high accuracy is a result of this non-linearity along with the availability of pre-trained word embeddings.
         \item 
-        
+        Multi-layer feed-forward networks can provide competitive results on sentiment classification and factoid question answering
+        \item 
+        Convolutional and pooling architecture show promising results on many tasks, including document classification, short-text categorization, sentiment classification, relation type classification between entities, event detection, paraphrase identification, semantic role labeling, question answering, predicting box-office revenues of movies based on critic reviews, modeling text interestingness, and modeling the relation between character-sequences and part-of-speech tags.
+        \item 
+        Convolutional and pooling architectures allow us to encode arbitrary large items as fixed size vectors capturing their most salient features, they do so by sacrificing most of the structural information.
+        \item 
+        Recurrent and recursive networks allows using sequences and trees and preserve the structural information.
     \end{itemize}
 
 % section a_primer_of_neural_net_models_for_nlp (end)