Basic Hebbian Learning
Brief description and instructions (DRAFT):
Hebb (1949) proposed a rule where by the synaptic strength between neurons would be increased. The rule can be mathematically expressed as follows:
Δw = αio
w is the strength of the connection between the presynaptic or input (i) and the postsynaptic or output (o) neuron.
Δw is the change in the strength.
α is the learning rate.
i is the activity of the input neuron
o is the activity of the output neuron.
Since the equation is multiplicative, it takes activity on both the input and output neuron for there to be any change in the learning strength. This applet will allow you to manipulate values to see this learning rule in action.
Using the illustration:
To keep things simple, there are only three neurons, two input neurons (A and B) and one output neuron. The current connection strength is indicate at the synapse between each input neuron and the output at the synapse. You can select the learning rule you want to run with the menu at top. You can chose a non-learning rule, a basic Hebb rule, and a simple competitive learning where the synapse that changes the most has an advantage and weakens synapses that do not change and reduces the learning of synapses that do change.
Below this menu you can start the animation and reset the learning to the beginning synaptic strengths. Below these buttons you can select the number of action potentials that each neuron will fire, if selected.
For each neuron, you can adjust the size of their release of neurotransmitter with the slider under the letter that has the neurons name. You can also chose if the neuron will fire and adjust the initial neuron synaptic strength that the reset button will reset the synaptic strength to.
Click here to open the applet. It will open a new window that will fill your screen.
Hebb, D. O. (1949). The Organization of Behavior. New York: Wiley.