This was my project for a hackathon that focuses on the usage of Data Analytics and Machine Learning.
The data was found from : http://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients#
Bank Evie is a well established bank who is recently experiencing a decline in their revenue and managed to trace
the source towards loan defaults. More people can’t afford to pay and consequently causing the bank to suffer the loses.
Aim of the project is to:
- Determine the common traits among these loan defaulters using Pandas & Numpy
- Build a model to predict the future defaulters based on their traits using Machine Learning
View the full presentation here: https://github.com/Viggy1357/datasciencehackathonBankEvie/blob/main/YME_Hackathon.ipynb It shares the insights from the analysis and next steps.
View the full codes here: https://github.com/Viggy1357/datasciencehackathonBankEvie/blob/main/YME_Hackathon.ipynb I used 4 different ML models to find the highest accuracy: Linear Regressions, SVM, Bayes and Stacked(Mixed)