Chapter 9: Other Popular XAI Frameworks
In the previous chapter, we covered the TCAV framework from Google AI, which is used for producing human-friendly concept-based explanations. We also discussed the other widely used explanation frameworks: LIME and SHAP. However, LIME, SHAP, and even TCAV have certain limitations, which we discussed in earlier chapters. None of these frameworks covers all the four dimensions of explainability for non-technical end-users. Due to these known drawbacks, the search for a robust Explainable AI (XAI) framework is still on.
The journey toward finding a robust XAI framework and addressing the known limitations of the popular XAI modules has led to the discovery and development of many other robust frameworks trying to address different aspects of ML model explainability. In this chapter, we will cover these other popular XAI frameworks apart from LIME, SHAP and TCAV.
More specifically, we will discuss about the important features, and key advantages...