Skip to content

3. Command line options

Jump to bottom
Ioannis G. Tsoulos edited this page Jun 2, 2018 · 1 revision

The available command line options of the main executable NNC are:

  1. -h. This options prints an information help screen and the program terminates.

  2. -p filename. This parameter determines the path of the objective problem. If the value of the parameter kind is neural, then this parameter determines the train file for classification or regression. Otherwise, this parameter determines the shared library with the differential equation to be solved.

  3. -t filename. The string parameter filename defines the test file for the neural network. This parameter is valid only if kind has the value of neural.

  4. -c count. The integer value count specifies the amount of chromosomes used in the genetic population. The default value for this parameter is 500.

  5. -l length. The integer parameter length determines the size of each chromosome. The default value is 200.

  6. -s rate. The float parameter rate specifies the selection rate used in the genetic algorithm. The default value is 0.1 (10%)

  7. -m rate. The float parameter rate stands for the mutation rate in the genetic algorithm. The default value is 0.05 (5%)

  8. -r seed. The integer parameter seed determines the number used as a random seed in random generators. The default value is 1

  9. -k kind. The string parameter kind determines the usage of the constructed neural network. Accepted values are

(a) neural, for regression or classification problems.

(b) ode, for solving ordinary differential equations.

(c) pde, for solving elliptical partial differential equations.

(d) sode, for solving systems of ordinary differential equations.

  1. -o filename. The string parameter filename sets the output file of the execution. The format of the output file depends on the parameter kind.

  2. -n gens. The integer parameter gens determines the maximum allowed number of generations. The default value is 500.

  3. -g gens. The integer parameter gens specifies the amount of generations that will be executed before the local search step of the genetic algorithm. The default value for this parameter is 50.

  4. -d count. The integer parameter count specifies the amount of chromosomes that will take part into local search step of the genetic algorithm. The default value for this parameter is 20.

Clone this wiki locally