TODO
Inspiration
The case was very intriguing from the technical aspect which drove our team members to think about solutions and innovate with new technologies such as artificial intelligence.
What it does
It analyses and gives an estimate of the success potential of a movie.
How we built it
It is built around a random forest classifier with an AUC equal to 97% on the test set. We coupled it with an interactive user web experience.
Challenges we ran into
We got tired during the night. We rely on a very limited data set that in the long term should be extended in order to make better and more relevant predictions.
Accomplishments that we're proud of
We created a basic functioning machine-learning model that predicts movie success with a high accuracy.
What we learned
We gained varied experience with different aspects of the development of a start-up application. We learned to work together with teammates that come from very diverse backgrounds.
What's next for PETSHARK!
We want to build a more sophisticated data infrastructure to be able to offer more complex analysis and predictions, for example location-based or time-based predictions using multiple variables.
Log in or sign up for Devpost to join the conversation.