Biomedical Engineering Reference
In-Depth Information
Fig. 1.
Schematic of a Neuron
intractable, models such as the Fitzhugh-Nagumo model have been cre-
ated as qualitative models of dendritic-somatic-axonal interactions 11 . An
artificial neuron is a minimalist qualitative model that suggests that
a neuron integrates variations in potential over time and res if a thresh-
hold is exceeded. An Artificial Neural Network (ANN) is a network of
interconnected articial neurons.
Specically, in an ANN, axonal action potentials are represented by
activations which are considered to be in [0; 1], and synaptic activity is
considered to be governed by a parameter known as a synaptic weight. If x j
denotes the activation from the j th neuron in a network, and if w ij is the
synaptic weight of the connection between the i th and j th articial neuron,
then the activation of the i th neuron is
0
@ X
j6=i
1
A
x i =
w ij x j i
(1)
where the activation function () is of the form
1
1 + e t
(t) =
where > 0 is a parameter 12 .
Fig. 2.
An Articial Neuron
Search WWH ::




Custom Search