Next: Population Size
Up: Parameter Selection
Previous: Parameter Selection
For the present application I use binary encoding of the 99 model
parameters (not to be confused with the evolution parameters that
control the inner workings of the genetic algorithm)
representing potential solutions to the problem ().
I represent a model parameter with 10 bits so that there are 210=1024
possible values for each model parameter and a total of 102499
possibilities for the entire search space. A completely exhaustive search
of the model space would therefore be very difficult if not impossible.
Next: Population Size
Up: Parameter Selection
Previous: Parameter Selection
Stanford Exploration Project
11/11/2002