Balanced Data Generator for Ethical AI
6.5
A Python library extension leveraging Mimesis to automate the generation of balanced, counterfactual datasets specifically for auditing and mitigating bias in machine learning models. Supplying a model with these datasets before training or for testing bias can significantly improve the ethical output of an AI.
80h
mvp estimate
6.5
viability grade
1
views
technology stack
Python
Easy
inspired by
Libraries can balance datasets to reduce model bias.