Self-optimised cost-sensitive classifiers for early field failure prediction in storage systems.
Mohamed Bader-El-DenTodd PerryPublished in: Swarm Evol. Comput. (2023)
Keyphrases
- cost sensitive
- failure prediction
- storage systems
- naive bayes
- boosting algorithms
- multi class
- cost sensitive learning
- misclassification costs
- cost sensitive classification
- binary classification
- decision trees
- file system
- training data
- support vector
- binary classifiers
- text classification
- classification accuracy
- feature selection
- class distribution
- class imbalance
- linear classifiers
- active learning
- support vector machine
- high performance computing
- training set
- databases
- test set
- flash memory
- learning process
- base classifiers
- class labels