Spaced Review Prediction for Language Learning
My team and I used data from a language learning app called Embark to create a model to predict the difficulty of a question and how likely it is for a given user to answer the question correctly. Because this involved sequential data with unknown parameters and observed parameters, we used several variations of Hidden Markov Models in our analysis.
In this project, I was part of the brainstorming, design, and troubleshooting of our algorithm. I wrote sections of the code and performed tests on my team member's code to ensure quality and functionality. I also wrote sections of the final paper and helped with revising and editing.