Ojas Rule
#phd #hebbian
2024-02-07
An addition to the basic hebbian learning rule that adds in the y to the subtracted term. List of Hebbian Weight Update Rules
When learning rates are low enough, it sorta approximates a normalization of the weights, so that the weights don’t grow out of control. Survey-of-Hebbian-Learning-Lagani-2023-CF4RA22D
If you want to align towards multiple principal components, you can do that, but you’ll need to include more than just one neuron.