Dynamic Pattern Selection for Faster Learning and Controlled Generalization of Neural Networks
by Axel Robel
http://www.ircam.fr/anasyn/roebel/paper/esann.ps.gz
Add To MetaCart
Abstract:
We address the question of selecting a proper training set for neural network time series prediction or function approximation. As a result of analyzing the relation between approximation and generalization, a new measure, the generalization factor, is introduced. Using this factor and cross validation we develop the dynamic pattern selection algorithm. By employing two time series prediction tasks, we compare the results for dynamic pattern selection training to results obtained with fixed training sets. The favorable properties of the dynamic pattern selection, namely lower computational expense and control of generalization, are demonstrated.
Citations
| 137 | Nonlinear Signal Processing Using Neural Networks – Lapedes, Farber - 1987 |
| 53 | How neural nets work – Lapedes, Farber - 1987 |
| 46 | Note on learning rate schedule for stochastic optimization – Darken, Moody - 1991 |
| 44 | Selecting concise training sets from clean data – Plutowski, White - 1993 |
| 8 | The Dynamic Pattern Selection Algorithm: Effective Training and Controlled Generalization of Backpropagation Neural Networks – Robel - 1994 |
| 3 | Estimation of the Kolmogerov entropy from chaotic signal – Grassberger, Procaccia - 1983 |
| 3 | Neural Models of Nonlinear Dynamical Systems and their Application to Musical Signals – Robel - 1993 |
| 1 | Dynamic selection of training patterns for neural networks: A new method to control the generalization – Robel - 1992 |
| 1 | Improving connectionsts learning, based on gradient descend – Salomon - 1991 |
