Predicting Credit Card Attrition

Inspiration

Credit card is important for banks to attract customers and make a profit. However, with fierce competition between banks, customers are not necessarily loyal to the service of one bank. If the credit card attrition can be predicted, the bank can improve its services and strategy to lower the attrition rate or spend more resources on potentially loyal customers.

what we do?

How we built it

Based on previous customers' demographic and behavior patterns of using their credit cards. We build a model that can predict customer attrition with an accuracy of around 75%.

Challenges we ran into

The most challenge we have is that the online environment is difficult for collaboration. Without the group dynamic, it was harder to communicate and brainstorm together.

Accomplishments that we're proud of

We are new to coding and barely know anything about machine learning, but we still manage to complete a project within a short period of time. And we are proud that we can create a project with only little knowledge about computer science.

What we learned

We learn that a meaningful project doesn't have to be complicated, and we can create something as long as we start to do it.

What's next for Predicting Credit Card Attrition

Perhaps making a recipe producer app or a website comparing prices of products, which are our ambition at the beginning of ideation.

Built With

• googlecolab
• googlesearch
• kaggle
• workshopslides