Capital One: Fraud Detection

Inspiration

Interest in machine learning algorithm and financial data

What it does

Detects fraudulent transaction

How we built it

python, pandas, sklearn, streamlit

Challenges we ran into

Converting into OOB, creating aesthetic UI

Accomplishments that we're proud of

Implementing multiple models and a user interface. Having an accuracy of 97% and an F1 score of roughly 0.9

What we learned

Machine learning algorithms, pandas, sklearn, data cleaning, streamlit, plotly

What's next for Capital One: Fraud Detection

Implementing OOB programming and polishing the UI, adding more graphs and allowing for more usability.

Built With

• python