Deterministic Neural Classification
- 1 June 2008
- journal article
- Published by MIT Press in Neural Computation
- Vol. 20 (6) , 1565-1595
- https://doi.org/10.1162/neco.2007.04-07-508
Abstract
This letter presents a minimum classification error learning formulation for a single-layer feedforward network (SLFN). By approximating the nonlinear counting step function using a quadratic function, the classification error rate is shown to be deterministically solvable. Essentially the derived solution is related to an existing weighted least-squares method with class-specific weights set according to the size of data set. By considering the class-specific weights as adjustable parameters, the learning formulation extends the classification robustness of the SLFN without sacrificing its intrinsic advantage of being a closed-form algorithm. While the method is applicable to other linear formulations, our empirical results indicate SLFN's effectiveness on classification generalization.Keywords
This publication has 33 references indexed in Scilit:
- Benchmarking a reduced multivariate polynomial pattern classifierPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2004
- Neural discriminant analysisIEEE Transactions on Neural Networks, 2000
- Linear classifiers by window trainingIEEE Transactions on Systems, Man, and Cybernetics, 1995
- On the discriminatory power of adaptive feed-forward layered networksPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1994
- Computing second derivatives in feed-forward networks: a reviewIEEE Transactions on Neural Networks, 1994
- Training feedforward networks with the Marquardt algorithmIEEE Transactions on Neural Networks, 1994
- Learning in neural networks by using tangent planes to constraint surfacesNeural Networks, 1993
- Discriminative learning for minimum error classification (pattern recognition)IEEE Transactions on Signal Processing, 1992
- First- and Second-Order Methods for Learning: Between Steepest Descent and Newton's MethodNeural Computation, 1992
- Optimization for training neural netsIEEE Transactions on Neural Networks, 1992