Inspiration
Attempted to use techniques that are on the DL side, more than traditional ML techniques
What it does
Created clusters based on unstructured text with medical references
How we built it
Used LSTM and tSNE, then LDA to obtain most important terms
Challenges we ran into
Building a working LSTM network, treating the original text, choosing number of words to consider, difficulty to clearly understand the cases
Accomplishments that we're proud of
Clusters were created based on data without labels
What we learned
Combining DL with tSNE and LDA later
Log in or sign up for Devpost to join the conversation.