Week 8 Homework

Resources

• Markov chain notes and videos
• Markov code examples
• Notes on LSTM
• p5-deeplearn-js LSTM example 1, p5-deeplearn-js LSTM example 2
• Notes on how to train your own LSTM for p5-deeplearn-js

Assignment

Use a markov chain or LSTM to generate text. It's not required to write new code for this assignment, you can use the examples verbatim but try new source texts and think about the context and design for the output. Here are some approaches you could take:

• Use one of the existing examples to generate text and pick different inputs. Play with the "order" and "maximum" length variables. Try mixing multiple texts. Copy paste your favorite outputs from the browser and document in a blog post.
• Use a markov chain on something other than text. Record your own sequence of daily habits. Try musical notes. Could pixel colors be evaluated as a markov chain? What else?
• Animate the process of the markov chain in action.
• Visualize n-gram frequencies.
• Create a bot that generates its output based on a markov chain.
• Use one of the pre-trained models to generate text with an LSTM and p5-deeplearn (pre-trained models coming soon!)
• Try researching a different machine learning framework (here are some options) and see if you can generate text with an RNN/LSTM.

Questions?

• add your question here

Homework Links

• name, links, etc.
• Koji -- Markov Chain Thesis question generator
• Nouf -
• Yeseul -- Pattern Generator using Markov Chain
• Richard -- LSTM trained on classifier
• Michael --
• Swapna -- recipe generator
• Melissa -- Jeopardy Generator first attempt
• Tong --
• Jenn --
• Laura -- Context-Free Counterpoint (Experimenting with last week's topic)
• Kenzo --
• Nitish --
• Paula -- [word searcher] (http://www.paulaleonvendagar.com/2017/10/31/markov-and-searcher-function/)
• Annie --
• Utsav --
• Zach --