Project Abstract

This research explores machine learning bias in gender, race, and age using the Adult dataset. As AI becomes more pervasive, it’s crucial to address biases for fairness. By analyzing logistic regression, decision trees, and SVM, this study aims to uncover and mitigate biases. Through data preprocessing, model training, and bias assessment, it reveals disparities in predictive outcomes. Ultimately, this research contributes to the dialogue on ethical AI development, advocating for fairness and transparency in AI systems. The project addresses several issues, including the serious effects that these biases are already having on some members of society, how we are responsible for AI’s biases and how we are ultimately responsible for finding a solution.

Keywords: AI Bias Research, Ethical human computer impact, Machine learning algorithm training

 Conference Details

Session: Poster Session B at Poster Stand 61

Location: Sir Stanley Clarke Auditorium at Wednesday 8th 09:00 – 12:30

Markers: Troy Astarte, Muneeb Ahmad

Course: BSc Computer Science, 3rd Year

Future Plans: I’m looking for work