Learning with limited numerical precision using the cascade-correlation algorithm.
Markus HöhfeldScott E. FahlmanPublished in: IEEE Trans. Neural Networks (1992)
Keyphrases
- learning algorithm
- experimental evaluation
- cascade correlation
- detection algorithm
- incremental learning
- dynamic programming
- learning process
- high accuracy
- learning speed
- matching algorithm
- preprocessing
- recurrent neural networks
- optimization algorithm
- worst case
- computational cost
- learning systems
- semi supervised
- cost function
- reinforcement learning
- k means
- active learning
- expert systems
- neural network
- data mining
- learning phase
- learning rules
- hidden markov models
- objective function
- learning tasks
- optimal solution
- significant improvement
- search space
- data sets
- segmentation algorithm
- theoretical analysis
- expectation maximization
- online learning
- supervised learning