Probabilistic Artificial Intelligence Seminar

Nothern Kentucky University (2005)

These are materials I wrote and presented for this 2005 seminar, to about a half dozen NKU Mathematics and Computer Science faculty.

I've included a KYMAA talk connecting Dempster-Shafer belief functions to Möbius transforms (not the Möbius transformations in complex analysis!). I liked the idea of getting away from beliefs as mere probabilities, to confront agnosticism about outcomes directly. And it was surprising to follow it to something I could think of as integration on a lattice.

We had a guiding book we read: Halpern (2003), Reasoning About Uncertainty . I see a second edition was later released, in 2017.

• Two-Ace Puzzle & Variant
  
• Rationality axioms and Dutch-book ustification of probability
• Upper/lower probabilities vs. outer/inner measures
  
• Mobius Transforms (KYMAA talk)
  
• Roadmap after 5 meetings
• Conditionalization and Jeffrey's Rule: Example

Photo (K. Kirby): Just monkeys in fog, Bagan, Myanmar, 2016.