Two members of this team are first year students. We figured a tool to help incoming students be better prepared for university life would be useful
What it does
Incoming university students answer a series of "What would you do?" questions based on choices they may encounter in university. Based on the user's decision making behaviour, the project will try to predict aspects of their university experience (i.e. predict the quality of their social life, mental health, academics, extracurricular involvement, etc.).
How we built it
The Flask framework was used to create a web app that is hosted with Microsoft Azure. Android studio was used to develop a similar android application. At the root of it all, the Scikit-Learn library for python was used to classify decision making behaviours (k-means clustering was used for unsupervised machine learning) to be used in predicting university experiences.
Challenges we ran into
Time constraint and exhaustion was the largest factor we ran into.
Accomplishments that we're proud of
Figuring how to do k-means clustering and designing the "what would you do?" questions were quite tough, but we pulled through! In addition, we had excellent communication among team members which required a lot of effort on all parts to do.
What we learned
How to use unsupervised machine learning with categorical (i.e non-ordinal) inputs.
What's next for Crossroads
Lots of things planned! Refine UI/UX. Improve upon machine learning algorithm used to predict university experience based on decision making behaviour. Gathering a large dataset. And most importantly, getting this project out there!