Skip to content

Repository files navigation

Reading of the data set. Removing the features that we don't required and Visualizing the data. Splitting the data into train test split. Applying one hot encoding. Making that data set having equal number of outputs. Making the data set balance. Generating Pkl files from the data. Training the model XG boost on the data. Making a web app for it.(Streamlit)

About

Detects fraudulent transactions in real time with 99.1% ROC-AUC across 284,000+ transactions, even though fraud makes up less than 0.2% of all activity. Built with an ensemble classifier and SMOTE for class imbalance.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages