Know Thy Model: Mitigating AI Errors [Virtual Sample Class]

Imagine you've developed an AI model to predict housing prices, and your boss wants to know how trustworthy it is. In particular, they want to know how likely it is that the model will be off by more than $10K. How would you determine the answer? This class will teach you how to quantify the reliability of your model in this way - not only for house price prediction, but for any high-stakes application area where AI mistakes can be costly. In general, we will discuss how much you should trust AI models given that no model is perfect. The professor will be Dr. Garrett Katz, Associate Professor in the Department of Electrical Engineering and Computer Science.