Skip to content

Latest commit

 

History

History
6 lines (4 loc) · 530 Bytes

README.md

File metadata and controls

6 lines (4 loc) · 530 Bytes

A guide to model calibration

A simple implementation of two methods of calibrating machine learning models - Platt scaling and isotonic regression. For the main notebook with a demo of the calibration on MNIST dataset, see Notebook.ipynb. The modules used are contained in model.py, calibration.py, and plotting.py.

For a short description of fitting a logistic regression line using linear regression, see Sigmoid regression.ipynb.