Some simple questions to test your knowledge of Threshold Logic units.
| Questions | Answers |
|---|---|
| What parts of the Simple Artificial Neuron correspond to the biological neuron? | Input Value = Action potential of afferent axon Input Lines = Dendrites Weights = Inhibitory or excitatory component of post-synaptic potential signal Summation Unit = Soma Threshold Unit = Axon hillock Output Lines = Efferent axons |
| What are the inputs lines? | The input lines represent the input values of the artificial neuron. The actual number of lines is variable, but their values are binary, an input line will have a value of either 0 or 1. |
| What are the weights? | The weights multiply the input value, by a positive or negative factor. This mimics the effect or excitatory or inhibitory actions in the biological neuron. A positive weight will model an excitatory action, a negative weight models an inhibitory action. |
| What is the summation unit? | The summation unit totals the input value and weight products of each input line to give a single value result. |
| What is the threshold unit? | The threshold unit is used to compare a given threshold value against the summation total. If the threshold value is exceeded or equalled then the output generated is 1. If the threshold is not reached the output is 0. |
| What are the output lines? | The output lines convey the output value away from the artifical neuron, either as an input to further artificial neuron nodes, or as a final result. |
| How and what does a synapse fire? | A synapse fires when action potential reaches an afferent axon terminal. At the pre-synaptic membrane, a neurotransmitter is released across the synaptic cleft, which chemically binds to the receptors at the post-synaptic membrane. The result is a post-synaptic potential (PSP) that spreads down from the synapse, down the dendritic tree to the soma. PSPs may be continuous and may be excitatory or inhibitory. |
| How and what does a simple artificial neuron input fire? | A input in the simple artificial neuron fires simply by mimicking the action potential, which is either present or not. The input fires if its value before the weight is applied is 1. |