PENCARIAN STOKASTIK HEURISTIK PADA ALGORITMA GENETIK UNTUK OPTIMASI PORTOFOLIO

Investment in shares is one of many kinds of efforts which is sensitive to

losses and profits. Due to its speculative nature, the prospects of profit will

depend very much on luck factor. One way to suppress the risk and maximize

profit is by share portfolio. Share portfolio classifies capitals into shares based on

the profit previously earned.

In order to avoid big losses, investors must be keen in choosing the right

shares and the proportion of shares. This can be done by a search process to arrive

at an optimal result starting with an estimate and rechecking before the decision is

made.

Genetic Algorithm is a biologically inspired stochastic and heuristics

maximization process that randomly selects two potential solutions from a

population of potential solutions. The process of Genetic Algorithm is based on

the natural selection mechanism and biological genetics to determine the structure

of high quality individuals in one population.

Each problem solving by Genetic Algorithm should be efficient and

effective, especially in making chromosome representation to the problem and

also in determining the code type related to the choice of elementary operator, that

is reproduction, crossover and mutation. Determination of a good fitness function

will have an effect on the solution to be obtained.