# Changes between Version 7 and Version 8 of Features/Genetic Programming

Ignore:
Timestamp:
01/06/12 00:13:20 (8 years ago)
Comment:

--

### Legend:

Unmodified
 v7 [[http://dev.heuristiclab.com/trac/hl/core/wiki/UsersSamples#GeneticProgramming | Samples for Genetic Programming]] HeuristicLab supports tree-based (Koza-style) genetic programming. This classical form of genetic programming uses a tree data structure as representation of solution candidates. The data types and operators to work with tree-based solution candidates are implemented in the plugin HeuristicLab.Encodings.SymbolicExpressionTree. !HeuristicLab supports tree-based (Koza-style) genetic programming. This classical form of genetic programming uses a tree data structure as representation of solution candidates. The data types and operators to work with tree-based solution candidates are implemented in the plugin HeuristicLab.Encodings.SymbolicExpressionTree. Supported operators (in 3.3.6): * Artificial Ant problem (HeuristicLab.Problems.ArtificialAnt) Samples for all three of these problems are available on the start page in HeuristicLab. Additional samples are available here: [[http://dev.heuristiclab.com/trac/hl/core/wiki/UsersSamples#GeneticProgramming]] Samples for all three of these problems are available on the start page in !HeuristicLab. Additional samples are available here: [[http://dev.heuristiclab.com/trac/hl/core/wiki/UsersSamples#GeneticProgramming]] == Symbolic regression == The symbols that can be used in symbolic regression models is problem dependent. So these symbols (mathematical operators: +,-,/,* ... and operands: constants, variables) are implemented in the problem specific plugin HeuristicLab.Problems.DataAnalysis.Symbolic. Evaluators for these problem-specific instances of symbolic expression trees are also implemented in this plugin. HeuristicLab provides many possibilities to analyse the resulting symbolic regression solutions. Some of them are shown in the screenshot. !HeuristicLab provides many possibilities to analyse the resulting symbolic regression solutions. Some of them are shown in the screenshot. Symbolic regression features: == Algorithms for Genetic Programming == Because of the strict separation of the implementations of algorithms, problems, and encodings in HeuristicLab. It is possible to solve symbolic regression of symbolic classification problems with all population based algorithms using either a crossover or manipulation operator to evolve solutions. The algorithm that comes closest to the original formulation of genetic programming by Koza is the genetic algorithm. Because of the strict separation of the implementations of algorithms, problems, and encodings in !HeuristicLab. It is possible to solve symbolic regression of symbolic classification problems with all population based algorithms using either a crossover or manipulation operator to evolve solutions. The algorithm that comes closest to the original formulation of genetic programming by Koza is the genetic algorithm. Available algorithms for genetic programming: