Rore: robust and efficient antioxidant protein classification via a novel dimensionality reduction strategy based on learning of fewer features

HIGHLIGHTS

• What: To address this problem, the authors propose the Rore model, which achieves superior MCC values based on a feature dimensionality reduction algorithm and also preserves the original information and relationships between the features. 2.3 Variational feature compressor To make the variational information bottleneck (VIB) available for improved feature dimensionality reduction, the authors propose a variational feature compressor (VFC). 3.2 Performance evaluation based on k‑fold cross‑validation To estimate the model performance more reliably, the k-fold cross-validation method was used to estimate model performance. The results of the experiments objectively demonstrated the . . .
  

   

  

  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.