Math
Genetic algorithm
Experiment with selection, crossover, mutation and elitism in a working genetic algorithm on multimodal fitness landscapes.
Try this
A population of 48 sixteen-bit chromosomes searches a landscape. Tournament selection, one-point crossover and per-bit mutation generate each generation. Elitism retains the best chromosome; global optimality is not guaranteed.
Equations used in the module
Equations
x = −3 + 6g/65535, g ∈ {0,…,65535}
Selection: best of three random individuals
Crossover: one random split between the 16 bits
Each bit flips independently with probability pmutation
