Reducing bias and ensuring fairness in data science

By Henry Hinnefeld

Civis Analytics
Feb 23, 2018 · 5 min read

Defining “Fairness”

Group vs. Individual Fairness

Balanced vs. Imbalanced Ground Truth

Sample Bias vs. Label Bias in your Data


Recommendations for Data Scientists

Think about the ground truth you are trying to model

Think about the process that generated your data

Keep a human in the loop, if your model affects people’s lives.

The Civis Journal

Civis Analytics helps the country's largest companies and nonprofits identify, attract, and engage loyal customers and employees with a blend of proprietary data, software solutions, and an interdisciplinary team of data and survey science experts.

Civis Analytics

Written by

Building a Data-Driven World | civisanalytics.com

The Civis Journal

Civis Analytics helps the country's largest companies and nonprofits identify, attract, and engage loyal customers and employees with a blend of proprietary data, software solutions, and an interdisciplinary team of data and survey science experts.