December 2023
Intermediate to advanced
320 pages
8h 29m
English
This chapter discusses some of the principle-specific techniques to address the RAI concerns around AI models and data. Among the ethical principles, security, reliability, and safety are considered traditional software quality attributes and well studied in the dependability community. Transparency, contestability, accountability, human-centered values, and human, societal, and environmental well-being are the principles that can be addressed by the governance patterns, process patterns, and product patterns introduced in the previous chapters. This chapter mainly focuses on the techniques to deal with fairness, privacy, and explainability, which are emerging quality attributes for AI systems. ...
Read now
Unlock full access