Compensated integrated gradients for reliable explanation of electroencephalogram signal classification

HIGHLIGHTS

• who: Yuji Kawai et al. from the Graduate School Engineering, Osaka University, Osaka, Japan have published the research work: Compensated Integrated Gradients for Reliable Explanation of Electroencephalogram Signal Classification, in the Journal: (JOURNAL) of 25/06/2022
• what: The authors propose a IG method that does not require a baseline which compensates the contributions calculated using the IG method at an arbitrary baseline by using an example the Shapley sampling value. The authors proposed a compensated IG method using one example of the SS value to improve the reliability of learning models in explaining classification . . .
  

   

  

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

  Thanks :)

  Username or E-mail:

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