Training a network with ternary weights using the CHIR algorithm.
Shai AbramsonDavid SaadEmanuel MaromPublished in: IEEE Trans. Neural Networks (1993)
Keyphrases
- learning algorithm
- experimental evaluation
- significant improvement
- dynamic programming
- times faster
- computational complexity
- detection algorithm
- high accuracy
- matching algorithm
- segmentation algorithm
- expectation maximization
- preprocessing
- search space
- cost function
- weight assignment
- weighted average
- convergence rate
- objective function
- optimal solution
- optimization algorithm
- worst case
- k means
- evolutionary algorithm
- input data
- neural network
- computational cost
- recognition algorithm
- training algorithm
- feed forward neural networks
- hidden nodes
- wireless sensor networks
- relative frequencies