- Predict the claim amount for MSA having less than 10 procedures.
- Validate the results for the DataRobot models developed by the J&J team.
- Develop a production level Machine Learning Framework for predicting Insurance Payments for surgical procedures.
- If possible, improve the model performance from the Baseline DataRobot models.
- Models and code to be delivered in Python.
Please run the following jupyter notebooks in sequence:
- Run the
Final Code Snippet/Capstone_Step1_Data_Preprocessing.ipynb
notebook responsible for transforming raw datasets into a preprocessed dataset. Please edit the file paths todata/raw
folder. - Afterwards, run
Final Code Snippet/Capstone_Step2_Data_Modelling.ipynb
responsible for performing data modelling and model evaluation. - Run
Final Code Snippet/Capstone_Step3_XGBoost_Model.ipynb
responsible for XGBoost modelling and enforcing the monotonicity constraints. Please ensure that output of Step 1 and Step 2 has been generated before running Step 3.
- Rahulraj Singh - Team Captain [
[email protected]
] - Prerit Jain [
[email protected]
] - Mahesh Jindal [
[email protected]
] - Parth Gupta [
[email protected]
] - Ayush Baral [
[email protected]
]
- Cindy Tong - Project Lead and Domain Expert
- Ziyu Tan - Technical Expert and go-to person for DataRobot
- Katherine Etter - Mentor and Knowledge Expert
Adam S. Kelleher [[email protected]
]
Xuanyu Li [[email protected]
]