Week 8 Homework

Machine Learning: RNN and Word2Vec

It is not required to write any new code for this assignment. You can simply run one or more of the examples and document your results in a blog post.

• Resources and examples for RNN
• Resources and examples for Word2Vec

RNN Ideas

• Can you train a model off of text from one of last week's examples or assignments? Compare and contrast the results. Here are instructions for training, a video you can follow, and a video about setting up a python environment.
• Consider altering the interaction of the generator. Can you invent ways for a writer to collaborate with the generator?

Word2Vec Ideas

• Find words along the path between two words using average().
• Compute the average vector for a block of text (sentence, paragraph, longer text). Then display the words closest to that average vector. Is this a way to do generate a keyword or tag for an article?
• Experiment with poetry generation using word vectors and average sentence vectors as in Allison Parrish's tutorial.
• Try training a word vector model with your own data, see instructions.

Name and URL

• Jenna Xu: mood predictor
• Yeonhee Lee - Critics
• Mengzhen - Taylor's Next Song
• Shawn - ELIZA - Chinese Poems Generator
• Zahra - LSTM Text Generation
• Itay - Storyteller - Creating a short story with a neural network trained on children stories.
• Chuyi - Portfolio Website Generator
• James - Headline Generator