insolver 0.4.9b0

Insolver is low-code machine learning library, initially created for the insurance industry, but can be used in any other. You can find a detailed overview at https://insolver.readthedocs.io/en/latest/source/overview.html.

Insolver

Insolver is low-code machine learning library, initially created for the insurance industry, but can be used in any other. A more detailed overview you can find here.

Installation:

pip install insolver

Install with addons:

pip install insolver[db-connects]

Post-install:

To fix displaying plotly figs in jyputerlab install:

jupyter labextension install jupyterlab-plotly

In case of problem with pyodbc you may need to install:

sudo apt install unixodbc-dev

Insolver is already installed in the easy access cloud via GitHub login. Try https://mset.space with a familiar notebook-style environment.

Examples:

• Private motor French insurer dataset
• US traffic accident dataset
• Landing club dataset

Documentation:

Available here

Supported libraries:

Libs:

• sklearn
• H2O

Boosting models:

• XGBoost
• LightGBM
• CatBoost

Model interpretation:

• shap plots

Serving (REST-API):

• flask
• fastapi

Run tests:

python -m pytest

tests with coverage:

python -m pytest --cov=insolver; coverage html; xdg-open htmlcov/index.html

Contributing to Insolver:

Please, feel free to open an issue or/and suggest PR, if you find any bugs or any enhancements.

Demo

Example of models creation using insolver

Example of model production service

Example of elyra pipeline built with insolver inside