Multiple instance neural networks based on sparse attention for cancer detection using t-cell receptor sequences

HIGHLIGHTS

• who: Younghoon Kim from the Department of Industrial, Medical Center, Dallas, TX, USA have published the Article: Multiple instance neural networks based on sparse attention for cancer detection using T-cell receptor sequences, in the Journal: (JOURNAL)
• what: The authors propose multiple instance neural networks based on sparse attention (MINN-SA) to enhance the performance in cancer detection and explainability. The experiments show that MINN-SA yields the highest area under the ROC curve scores on average measured across 10 different types of cancers compared to existing MIL approaches. The authors propose a sparse attention . . .
  

   

  

  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.