Explaining a series of models by propagating shapley values

HIGHLIGHTS

• who: Hugh Chen from the University of Washington, Seattle, Microsoft Research, Redmond, WA, USA. have published the paper: Explaining a series of models by propagating Shapley values, in the Journal: NATURE COMMUNICATIONS NATURE COMMUNICATIONS
• what: The authors propose a theoretical framework (Methods section Connecting DeepLIFT`s rules to the Shapley values) that connects the rules introduced in ref 17 to the Shapley value with an interventional conditional expectation set function (with a flat causal graph; i.e., a causal graph where arrows are only drawn between input variables and the output) (ICE Shapley value) (Methods . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.