Development of Robust Adaptive Inverse models using Bacterial Foraging Optimization

Adaptive inverse models find applications in communication and magnetic channel equalization, recovery of digital data and adaptive linearization of sensor characteristics. In presence of outliers in the training signal, the model accuracy is severely reduced. In this paper three robust inverse models are developed by recursively minimizing robust norms using BFO based learning rule. The performance of these models is assesses through simulation study and is compared with those obtained by standard squared norm based models. It is in general, observed that the Wilcoxon norm based model provides best performance. Moreover the squared error based model is observed to perform the worst.