GA for single variable

Can I use genetic algorithm for single variable constrained optimization?

my function is f (x) = a/(x-log(x/t))

s.t. f(x) < 1

where a & t are constants

Answer 1

[Chunru]

Yes, you can. Howerver, no gurantee of the optimal solution (just like any other non-linear optimization approaches).

a = 1;
t = 1;
xopt = ga(@(x) min(a./(x-log(x/t)), 1), 1)

Optimization terminated: stall generations limit exceeded.

xopt = -555.4795

a./(xopt-log(xopt/t))

ans = -0.0018 + 0.0000i

[Swapnil Kavitkar]

what about population, mutation ,crossover?? GA options

[Chunru]

"doc ga" to find out all the options available.