Build upon your foundational knowledge of natural language processing (NLP) by exploring more complex topics such as word2vec, doc2vec, and recurrent neural networks.
Overview
Syllabus
Introduction
- Leveraging the power of messy text data
- What you should know
- What tools you need
- Using the exercise files
- What is NLP?
- NLTK setup
- Reading text data into Python
- Cleaning text data
- Vectorize text using TF-IDF
- Building a model on top of vectorized text
- What is word2vec?
- What makes word2vec powerful?
- How to implement word2vec
- How to prep word vectors for modeling
- What is doc2vec?
- What makes doc2vec powerful?
- How to implement doc2vec
- How to prep document vectors for modeling
- What is a neural network?
- What is a recurrent neural network?
- What makes RNNs so powerful for NLP problems?
- Preparing data for an RNN
- How to implement a basic RNN
- Prep the data for modeling
- Build a model on TF-IDF vectors
- Build a model on word2vec embeddings
- Build a model on doc2vec embeddings
- Build an RNN model
- Compare all methods using key performance metrics
- Key takeaways for advanced NLP modeling techniques
- How to continue advancing your skills
Taught by
Derek Jedamski