This survey discusses Genetic Programming (GP) and its multi-dimensional applications. GP is viewed as one of the branches of evolutionary computation, since it has the ability to evolve executable computer programs. Further, it is a robust, flexible and intuitive artificial intelligence paradigm, having a wide range of applications. With the help of genetic programming, one can solve systems that interact with real world, make theories about consequences of their actions and dynamically adjust inductive bias. In addition, there are applications of genetic programming like solving tasks of data mining, robotic soccer gamewhere layered learning approach has been implemented, learning with imprecise or missing datawherein genetic programming approximates the missing values by using correlation input data and eliminating the need of pre-processing and financial genetic programming, etc.
Genetic
Programming (GP) introduced by John R Koza (1992) and his group is one of the
several problem-solving methods based on a computational analogy to natural evolution.
It is an area defined under the umbrella of evolutionary computation. Genetic
programming evolves executable computer program using the ideas of biological
evolution to handle a complex problem. Out of a number of possible programs, the
most effective programs survive and compete with other programs to continually
approach closer to the needed solution. This is an approach that seems most appropriate
with problems in which there are a large number of fluctuating variables.
There
are numerous applications of genetic programming including black-art-problems,
programming-the-un-programmable and many more. In this survey, we will discuss
mainly four applications. These are data mining tasks, robotic soccer game, learning
with imprecise or missing data, and financial genetic programming. Applications
of GP are increasing on new domain areas to be applicable to the techniques of
genetic algorithms for achieving human-competitive machine intelligence. |
