Tutorial 9: Algorithmic Bias, Accessibility, and Equity
Tutorial 9: Algorithmic Bias, Accessibility, and Equity#
In week 8, Marc introduced the concept of algorithmic bias, and emphasized the importance of creating accessible AI systems. Accessible systems should be as usable for those with disability as those without, considering both permanent and temporary disabilities, as well as situational impairments. The importance of equitable systems (which promote fairness, and are impartial) was also highlighted.
The aims of the tutorial on Accessibility and Equity are as follows.
To gain an awareness of how machine learning models can encode human bias.
To critically discuss a case study/thought experiment on the issues of accessibility and equity to an AI system once it is deployed.
No summary this week, lots of exercises instead!