Model Deployment for Credit Card Fraud Detection

This is the part that we deployed the best performance model we built from the ML Olympaid Data Hackton.
Because model deployment is not part of the competition. Hence, this github repo completes the whole cycle of data science project.

Use Case

This is the classfication use case to predict the credit card transaction validity. We are required to build the predictive model to acheive the objective. For the earlier about this project. Please visit the following [Github] (https://github.com/hoe94/ML-Olympaid-Data-Hackathon) to understand EDA, Feature Engineering & Model Creation part.
In this stage, we picked the best performance model and deploy into API with Flask microframework.
Lastly, we hosted the API on Heroku (PaaS). You may try this [API] (https://creditcard-frauddetection-api.herokuapp.com) with the input parameter & POST method.
Below GIF is the demo about how to run the API for the predictive modeling.