eduzhai > Applied Sciences > Engineering >

Host Load Prediction with Bi-directional Long Short-Term Memory in Cloud Computing

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 9 pages

Abstract: Host load prediction is the basic decision information for managing thecomputing resources usage on the cloud platform, its accuracy is critical forachieving the servicelevel agreement. Host load data in cloud environment ismore high volatility and noise compared to that of grid computing, traditionaldata-driven methods tend to have low predictive accuracy when dealing with hostload of cloud computing, Thus, we have proposed a host load prediction methodbased on Bidirectional Long Short-Term Memory (BiLSTM) in this paper. OurBiLSTM-based apporach improve the memory capbility and nonlinear modelingability of LSTM and LSTM Encoder-Decoder (LSTM-ED), which is used in the recentprevious work, In order to evaluate our approach, we have conducted experimentsusing a 1-month trace of a Google data centre with more than twelve thousandmachines. our BiLSTM-based approach successfully achieves higher accuracy thanother previous models, including the recent LSTM one and LSTM-ED one.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×