Skip to content

notes

Jump to bottom Edit New page
Kailuo Wang edited this page Mar 15, 2015 · 3 revisions
  • Is general user interaction a POMDP(partially observed Markov decision process)?
  • Might need to deduce user reaction to MDP
  • Is user interaction episodic? (each session as one episode, if so do we have enough episodes for training?)
    • MC can only be applied to episodic MDP
Add a custom footer
Add a custom sidebar
Clone this wiki locally