Metadata-Version: 2.1
Name: coarseclassevaluator
Version: 1.1
Summary: CoarseClassvisual formulator
Home-page: UNKNOWN
Author: Asheesh Janda
Author-email: asheesh.janda.ds@gmail.com
License: UNKNOWN
Project-URL: Source, https://github.com/AsheeshJandaDS/coarseclassvisual
Project-URL: Tracker, https://github.com/AsheeshJandaDS/coarseclassvisual/issues
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
License-File: LICENSE.txt

# Project Discription

### CoarseClassvisual:
This package helps a Data scientist to reduce the number of levels inside the category to <=5 levels for a binary classification problems, 
This package is the implementation of WOE based binning, and uses similar WOE to collate bin.

Below are the applications of this package

1. Explicability: Improves explicability of the story teeling of patterns. 
Instead of saying NY, SF, Seattle has more loan defaults this can logically group cities using this package and say all the hi-tech & Financial sectors are has more loan defaults
2. Reduces Noise and sparsity: With too many levels of predictor of binary classification , It becomes too noisy and upon dummy variables increases sparsity. This package helps
one reduce the Noise and sparsity


