Intelligent failure prediction models for scientific workflows

作者：

Highlights：

• Intelligent task failure models using machine learning approaches are proposed.

• The accuracy of proposed models is validated in Pegasus and Amazon EC2.

• The prediction accuracy of (94%) is achieved using Naïve Bayes approach.

摘要

•Intelligent task failure models using machine learning approaches are proposed.•The accuracy of proposed models is validated in Pegasus and Amazon EC2.•The prediction accuracy of (94%) is achieved using Naïve Bayes approach.

论文关键词：Cloud Computing,Workflows,Failure prediction,Scientific workflows,Machine learning

论文评审过程：Available online 21 September 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.09.014