cognitive psychology. PhD 
Time to live-tweet the probabilistic graphical models course I'm taking. Prepare for a quarter of Hot Takes on how probabilistic programming languages are the future. cc @stephtwang
Show this thread
-
Marginalization is such a powerful concept in probability. In decision-making, I think we humans tend to infer from context a few most likely possibilities for unobserved quantities. Marginalizing, by contrast, is like simultaneously imagining every possible state of the world.
Show this thread
While classic strategy algorithms like minimax employ similar kinds of search, they make assumptions like "I know how to play the game, and the opponent thinks like me." Marginalization (and expectation) allow inference of optimal strategy just based on previous observations.
-
-
Love it when physicists invent algorithms instead of mathematicians. They give names with wonderful analogies to the real world, like using "temperature" in simulated annealing as a metaphor for concentration of "energy".
Show this thread -
A good reminder that most statistical methods don't just arise out of a vacuum, but rather from people who had really concrete needs (simulating atomic bombs on extremely primitive hardware, etc.). Another data point in the "does important math come from mathematicians" debate!
Show this thread -
Summarized some thoughts from the quarter in "Compiling Knowledge into Probabilities." Why can't we design probabilistic programs the same way we design deterministic ones? http://willcrichton.net/notes/compiling-knowledge-probability/ …
Show this thread -
Thanks again @stephtwang (and other hardworking TAs) for a fun quarter!
Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.