Skip to content

Simulate bias in data, evaluate machine learning models trained on biased datasets, and assess various bias mitigation techniques to improve fairness in predictions. Bias creation, analysis, model training, and fairness evaluation.

Notifications You must be signed in to change notification settings

PRCSousa/Tabular-Bias-Simulation-And-Mitigation

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

31 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Tabular Data Bias Simulation, Analysis and Mitigaion

Project Overview

This project aims to simulate bias in data and evaluate the behavior of machine learning models trained on biased datasets. The project will further explore various methods for bias mitigation, assessing their effectiveness in improving the fairness of model predictions. By studying the impact of biased data on machine learning systems, this project helps to understand the challenges posed by bias and how to address them in a practical way.

Objectives

  1. Simulate Bias: Introduce controlled biases into datasets to study how they affect machine learning model training and performance.
  2. Train Models: Train machine learning models on these biased datasets and analyze their predictions.
  3. Assess Impact: Understand the influence of biased data on model behavior, performance, and the fairness of predictions.
  4. Mitigate Bias: Apply bias mitigation techniques to the trained models and evaluate their effectiveness in improving fairness.
  5. Fairness Evaluation: Use fairness metrics to assess how different interventions influence the fairness of the model's predictions.

About

Simulate bias in data, evaluate machine learning models trained on biased datasets, and assess various bias mitigation techniques to improve fairness in predictions. Bias creation, analysis, model training, and fairness evaluation.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published