by Antonio Carvajal-Rodríguez ().
Área de Genética, Departamento de Bioquímica, Genética e Inmunología. Universidad de Vigo, Spain.
A. Carvajal-Rodríguez 2004. GASP: Genetic Algorithm (based on) Surviving Probability. Online Journal of Bioinformatics 5: 23-31.
SGA Algorithm
1.- Generate randomly N individuals.
2.- Calculate the fitness of each individual in the population.
3.- Select randomly a pair of parents from current population, the probability of selection being an increasing function of fitness (fitness proportionate selection).
4.- Obtain a pair of sons by recombination from each pair of parents.
5.- Repeat step 3 until the population size N is reached.
6.- The new population substitutes the old one.
7.- Mutate and iterate using this new population.
GASP Algorithm
1.- Generate randomly Nini individuals
2.- Select randomly a pair of parents
3.- Obtain one son by recombination from each pair of parents . Add one more attempt to the particular counter of one progenitor.
4.- Evaluate the fitness of each son. The fitness is obtained dividing the objective function value by K which is the maximum or a reasonable supreme (higher than the maximum) of the objective function. A son will survive when its fitness value is higher than a number randomly generated from an uniform between 0 and 1.
5.- Repeat step 2 until the maximum population size is reached or all parents have spent their fixed number (R) of attempts.
6.- The new population substitutes the old one.
7.- Mutate and iterate using this new population.
GASP Executable
Linux (upon request)
DOS (upon request)
Code (in C) (upon request)