Comparing Logs and Models of Parallel Workloads Using the Co-plot Method.
David TalbyDror G. FeitelsonAdi RavehPublished in: JSSPP (1999)
Keyphrases
- high precision
- experimental evaluation
- preprocessing
- similarity measure
- statistical model
- machine learning methods
- computational complexity
- statistical models
- cost function
- segmentation method
- prediction model
- mathematical model
- detection method
- parameter estimation
- high accuracy
- probabilistic model
- computational cost
- significant improvement
- prior knowledge
- pairwise
- multiscale
- bayesian framework
- statistical methods
- linear model
- support vector machine svm
- model selection
- classification accuracy
- k means