Adaptation and Learning in Automatic Systems
Adaptation and Learning in Automatic Systems
Robust estimation via stochastic approximation
IEEE Transactions on Information Theory
Hi-index | 22.14 |
Convenient identification techniques based on maximum likelihood estimators (MLE) are very sensitive to deviations from assumed distributions of observations. Huber's approach to robust estimation is highly fruitful for solving identification problems under incomplete information. In the paper some robust estimators for nonlinear regression problems are proposed and their features are discussed.