Abstract | ||
---|---|---|
The approach is an extension of the method of radial basis functions. Parameter estimation for the nonlinear predictor is performed by a gradient descent over a mean squared error measure, starting from a random initialization of the parameters. Results on predicting segments of speech data and the sunspot series are presented and compared to a linear predictor. An approach to adaptive estimation of the model by means of an extended Kalman filter is presented. In terms of prediction residual, the nonlinear predictor is found to perform significantly better than a linear model with the same number of parameters. Difficulties in applying this model in speech processing are discussed. |
Year | DOI | Venue |
---|---|---|
1991 | 10.1109/ICASSP.1991.150640 | ICASSP '91 Proceedings of the Acoustics, Speech, and Signal Processing, 1991. ICASSP-91., 1991 International Conference |
Keywords | DocType | ISSN |
error measure,nonlinear model,parameter estimation,radial basis function,extended kalman filter,speech data,linear predictor,nonlinear predictor,signal interpolation,gradient descent,linear model,time series prediction,speech processing,radial basis functions,kalman filters,mean square error,nonlinear systems,interpolation,vectors,time series,least squares approximation,neural networks,context modeling,predictive models | Conference | 1520-6149 |
ISBN | Citations | PageRank |
0-7803-0003-3 | 8 | 3.97 |
References | Authors | |
2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mahesan Niranjan | 1 | 775 | 120.43 |
V. Kadirkamanathan | 2 | 355 | 39.25 |