AI Interpretability & Transparency Question 3
You're working on an image classification model for identifying different types of clouds. During testing, you notice some strange results. The model seems to be focusing on irrelevant areas of the images (like background objects) instead of the actual cloud features. To understand why your model is behaving this way, which of the following tools or techniques would be the most helpful?
本练习是课程的一部分
