Debugging toward Responsible AI
Developing successful machine learning models is not solely about achieving high performance. We all get excited when we improve the performance of our models. We feel responsible for developing a high-performance model. But we are also responsible for building fair and secure models. These goals, which are beyond performance improvement, are among the objectives of responsible machine learning, or more broadly, responsible artificial intelligence. As part of responsible machine learning modeling, we should consider transparency and accountability when training and making predictions for our models and consider governance systems for our data and modeling processes.
In this chapter, we will cover the following topics:
- Impartial modeling fairness in machine learning
- Security and privacy in machine learning
- Transparency in machine learning modeling
- Accountable and open to inspection modeling
- Data and model governance
By...
