Error-adaptive classifier boosting (EACB): Exploiting data-driven training for highly fault-tolerant hardware.
Zhuo WangRobert E. SchapireNaveen VermaPublished in: ICASSP (2014)
Keyphrases
- fault tolerant
- data driven
- training error
- fault tolerance
- boosted classifiers
- adaboost algorithm
- training process
- error detection
- training set
- training examples
- error rate
- discriminative classifiers
- distributed systems
- generalization error
- training samples
- weak learners
- learning algorithm
- feature selection
- training data
- ensemble learning
- weak classifiers
- low cost
- state machine
- high availability
- load balancing
- decision stumps
- support vector machine
- face detection
- ensemble classifier
- boosting algorithms
- supervised learning
- evolvable hardware
- decision trees
- active learning
- object detectors
- safety critical
- computer architecture
- feature set
- multiple classifier systems
- computer systems
- operating system