What is Evolutionary Computation?
Evolutionary computation, a growing area of computer science, finds its inspiration in nature. The biological mechanism of evolution is mirrored in software and applied to several classes of problems. There are many evolutionary algorithms including genetic algorithms, evolution strategies, genetic programming and classifier systems.
Methodology
The computer randomly generates an initial population of candidate solutions to a problem. From this group, some parents are selected to produce new candidate solutions or children using genetic operators. All members of the population are rated according to their ability to solve the given problem. This fitness value determines which members form the next generation.
Benefits
Many offspring are produced in each generation so the process is intrinsically parallel. Evolutionary algorithms are especially well-suited to solving problems where the landscape of potential solutions is massive. The genetic approach may be indicated in cases where it would take an inordinate amount of time to step through the entire solution space. Over dozens or hundreds of generations, a computer can efficiently find a solution to a complex problem in which many parameters must be adjusted simultaneously.
Evolutionary computation software enables computers to find answers in the same way that nature adapts living species.
Our Primary Product
Genetic System Search for Technical Analysis employs evolutionary computation to search for investment trading systems. Read more about the software at the GSSTA website.