Next: Intuition for Perceptron Training
Up: Neural Network Learning
Previous: Perceptron Learning Algorithms
- begin with random weights, modify them, repeat until the
perceptron classifies all training examples correctly
- perceptron training rule:
Бе, where 
Бе -
is the target output, 
is the output generated by the
perceptron, and 
is the learning rateБе which moderates the
degree to which weights are changed at each step, usually set to a
small value (0.1)
Patricia Riddle
Fri May 15 13:00:36 NZST 1998