Employing Decoding of Specific Error Correcting Codes as a New Classification Criterion in Multiclass Learning Problems.
Yurong LuoKayvan NajarianPublished in: ICPR (2010)
Keyphrases
- multi class
- learning problems
- binary classification
- error correcting output codes
- multiclass classification
- binary classifiers
- multiclass learning
- machine learning algorithms
- multiclass problems
- supervised learning
- support vector machine
- feature selection
- error correcting codes
- cost sensitive
- binary classification problems
- multi task
- learning tasks
- machine learning
- kernel methods
- learning algorithm
- reinforcement learning
- error correcting
- support vector
- decoding algorithm
- multiple kernel learning
- decision trees
- classification algorithm
- feature set
- pairwise
- support vector machine svm
- classification accuracy
- high dimensional
- feature vectors
- svm classifier
- feature space
- active learning
- multi task learning
- semi supervised learning
- image segmentation
- data sets
- feature extraction
- training data
- generalization ability
- base classifiers
- unsupervised learning
- text classification