Energy functions for minimizing misclassification error with minimum-complexity networks

Neural Networks

Volumn 7, Issue 5, 1994, Pages 809-817

  Telfer, Brian A ^a   Szu, Harold H ^a  

^a NAVAL SURFACE WARFARE CENTER   (United States)

Author keywords

Classification; Conjugate gradient; Energy functions; Gradient descent; Minimum misclassification error; Neural networks; Objective functions

Indexed keywords

APPROXIMATION THEORY; CLASSIFICATION (OF INFORMATION); COMPUTATIONAL COMPLEXITY; ERRORS; FUNCTION EVALUATION; LEAST SQUARES APPROXIMATIONS; PROBABILITY;

CONJUGATE GRADIENTS; GRADIENT DESCENT; MINIMUM MISCLASSIFICATION ERROR ENERGY FUNCTIONS; OBJECTIVE FUNCTIONS;

NEURAL NETWORKS;

EID: 0027940939     PISSN: 08936080     EISSN: None     Source Type: Journal    
DOI: 10.1016/0893-6080(94)90102-3     Document Type: Article

References (24)

1. Minimum complexity density estimation
2. Performance and generalization of the classification figure of merit criterion function
3. Optimization for training neural nets
4. A comparison between criterion functions for linear classifiers
5. Differential theory of learning for efficient neural network pattern recognition
6. Shooting craps in search of an optimal strategy for training connectionist pattern classifiers
7. Why error measures are sub-optimal for training neural network pattern classifiers
8. Equivalence proofs for multilayer perceptron classifiers and the Bayesian discriminant function
9. A novel objective function for improved phoneme recognition using time-delay neural networks
10. The design and analysis of pattern recognition experiments
11. The aspirin/migraines software tools user's manual release v5.0
12. Neural network classifiers estimate Bayesian a posteriori probabilities
13. Minimum misclassification error performance measure for layered networks of artificial neurons
14. Minimum-misclassification-error neural networks for target classification
15. Implementing the minimum-misclassification-error energy function for target recognition
16. Classifying multispectral data by neural networks
17. Training a one-dimensional classifier to minimize the probability of error
18. Beyond regression: New tools for prediction and analysis in behavioral sciences
19. 30 years of adaptive neural networks: Perceptron, madaline, and backpropagation