"Next Generation AI" Topic I: AI and Uncertainty
Decision making through AI-based algorithms can be affected detrimentally by unperceived or irrelevant changes of the input, e.g. in object recognition. Certified robustness of the decision, for example to natural variations, however, are of utmost importance for basically all applications. Thus, uncertainty in the input data, but also quantifying uncertainty of the output is currently a major concern. This is intimately connected to a more psychological question: If AI mimics human behavior, which type of and how many failures due to uncertainties do we accept of AI methods? In this workshop, such questions and current state-of-the-art approaches in different application areas will be presented and discussed.
This is the first of a series of four roundtables where LMU researchers will discuss the topic of "Next Generation AI".
Location and Registration
- Poster "Next Generation AI" Roundtables (428 KByte)
- Program "Next Generation AI" Topic I: AI and Uncertainty (279 KByte)