Open aabuyazid opened 7 months ago
In order for a perceptron to "learn", it will need to update its weights proportional its contribution to the error. Mean-Square Error (MSE) is a very good place to start, but we should be able to change the error function in the future.
In order for a perceptron to "learn", it will need to update its weights proportional its contribution to the error. Mean-Square Error (MSE) is a very good place to start, but we should be able to change the error function in the future.