From the course: Foundations of Responsible AI
Join today to access over 24,600 courses taught by industry experts.
Technical aspects of sociotechnical solutions
From the course: Foundations of Responsible AI
Technical aspects of sociotechnical solutions
- We've already discussed how important it is to have social and behavioral experts at the forefront of AI development. So, let's discuss some of the ways to address bias with statistical tools. Studies have shown that we can't create somewhat fair AI without investigating data sets that include sensitive features such as age, gender, and race. Because algorithms can still be biased even without being trained on sensitive features, we have to work towards mitigating bias. Non-sensitive features in our data set may be correlated with sensitive features. So, if we remove those features, we don't know our algorithms exhibit bias. In general, mitigating bias in the outcomes of machine learning systems is the goal of responsible AI. First, if we want to make bias systems better, we have to identify what kinds of bias they exhibit. Then take steps to mitigate those biases. For identifying bias in AI there are four common…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
