Michael95-m / packaging-insurance-claim-model

Packaging regression model from scikit-learn

Packaging insurance claim model

Steps I make through this repository

• notebook

  You can see the EDA, feature engineering and modeling notebook inside notebook folder.

• source code

  You can see the main source code that make from above notebooks in insurance_claim_model folder.

Actually I didn't give much time to EDA and feature engineering parts because I want to emphasize and focus to make the well-structured code with good practices like testing with tox and pytest and using lint tools.

• Packaging

  You can see MANIFEST.in, pyproject.toml and setup.py which are essential for python packaging.

The steps I make for this repo is that at first, I make EDA and modeling notebook. From these notebooks, I create the python files(.py). Then make the python package which can installed with pip. You can see my python package in here.

About Dataset

The dataset I use in this repository is from here. It is about insurance claim analysis from demographics and health factors. It is released on 2023.

I make a regression model with random forest which can predict insurance claim from that dataset.