Simulation of the Action Potential in the Neuron’s Membrane in Artificial Neural Networks - Advancing Artificial Intelligence Through Biological Process Applications - page 84

Figure 10. Genetic Algorithm parameters

Figure 11. Parameters of the ANN Architecture

Another problem to be solved will be the predic-
tion of a classical laboratory random time series:
the Mackey-Glass series (Mackey, 1977).
Most of these time series cases are structures
with one single input and one single output. The
input corresponds to a number value of the t
moment, while the system's output is the pre-

diction of the number value at t+1 (short term
predictions).
First, it's showed the parameters used in the
ANN developed to solve the first two problems
(sunspots and tobacco). These parameters mean:
Time (the scale of the line gradient), Iterations
(number of internal cycles of the ANN before it