The Multi-Class Imbalance Problem: Cost Functions with Modular and Non-Modular Neural Networks.
Roberto AlejoJosé Martínez SotocaRosa Maria ValdovinosGustavo A. CasañPublished in: ISNN (4) (2009)
Keyphrases
- modular neural networks
- class imbalance
- cost function
- evolutionary learning
- class distribution
- bio inspired
- active learning
- cost sensitive
- imbalanced datasets
- cost sensitive learning
- high dimensionality
- feature selection
- pattern recognition
- sampling methods
- concept drift
- minority class
- evolutionary computation
- hybrid intelligent systems
- training data
- training examples
- neural network
- computational intelligence
- feature vectors