Extended Kalman Filter for Training Neural-Networks
An implementation of chapter two from Kalman Filtering and Neural Networks by Haykin.
The file knn.py contains a class for creating and training a multilayer perceptron by either extended Kalman filter (Newton-Raphson) or stochastic gradient descent. Enjoy!
