README.md

Introduction to Deep Learning for Natural Language Processing (Twitter Data)

This repo accompanies the Introduction to Deep Learning for Natural Language Processing to explain the core concepts of deep learning with emphasis on classifying text as the application. Python is used

Overview

The following topics are covered

1. What is deep learning?
2. Motivation: Some use cases
3. Building blocks of Neural Networks (Neuron, Activation Function, Backpropagation Algorithm)
4. Word Embedding
5. word2vec
6. Introduction to keras
7. Multi-layer perceptron
8. Convolutional Neural Network
9. Recurrent Neural Network
10. Challenges in Deep Learning

Project

The first step is to obtain access and register the application to interact with the Twitter API. Once the application has been registered, tweets can be extracted using the tweepy package and the Twitter API. The extracted tweets will be stored in the JSON format. The tweets stored in the JSON format are accessed and and will be used as training data. Word2Vec and a LSTM model will be used to generate new tweets from the training data.

Slides

The slides used are available here