Inspiration
The deep blue sea movie had lots of people doing stupid things and dying, so we decided to see if the average human was as stupid
What it does
We made an application that will determine if you will devoured by a shark based off historic data and machine learning in the form of an interactive and fun website.
How we built it
We got our dataset from Kagle which is basically Shark attack data. Then we saw the attributes which would be needed to predict and dropped the rest of the columns. Now since most of our attributes are categorical data so we used One-Hot encoding (For: Sex and Time columns) and Target encoding (For: Country and activities). Now all of our data is in numeric form so then we used SMOTE to balance our data as we had a lot of None-fatal cases compared to fatal-cases. Then we finally used logistic regression to train our model and saved it as a pickle file (Check out the .ipy file in Machine learning directory for complete code).
Challenges we ran into
We ran into several problems related to flask and sending data from the HTML form to the classifier.
Accomplishments that we're proud of
We're happy that we integrated Flask with an ML model to accomplish an idea.
What we learned
We figured out how to utilize Machine Learning in a choice based game. We additionally figured out how to handle issues as we were acquainted with more newer issues at each step of the making of our game.
What's next for Sharkies
Improve UX and UI and increase accuracy of the model.
Log in or sign up for Devpost to join the conversation.