RiversoftAVG has a Genetic Algorithms & Programming Component Library for Delphi XE8 Firemonkey. It is a powerful genetic algorithms and genetic programming solution for Android, IOS, OSX, and Windows. We’ve covered it before for Delphi XE5. On the genetic algorithm side it includes six different selection methods which are Roulette, Random, Tournament, Stochastic Tournament, Elitist, or Custom. Four fitness search methods are also included which are Minimize, Maximize, Weighted Minimize, and Weighted Maximize. Three genetic operations including crossover, mutation, and inversion are supported. Five crossover methods which are included are gene boundary, bit boundary, byte boundary, word boundary, and long word boundary. On the genetic programming side it includes six initialization methods which are Full, Grow, Half and Half, Ramped Full, Ramped Grow, and Ramped Half and Half. Three basic genetic operations which are crossover, mutation, and inversion. Six different mutation methods which are Subtree, Replacement, Constant, Shrink, Hoist, and Point. And seventeen different bloat control strategies which are Limit Size or Depth, Tarpeian Size or Depth, Unfit Size or Depth, Shrink Size or Depth, Hoist Size or Depth, Size Fair or Depth Fair Crossover, Size or Depth Parsimony Pressure, Covariant Size or Depth Parsimony Pressure, and Lexicographic Parsimony Pressure. As you can see this is a massive commercial library which comes with full source code. There is a free trial available for Delphi and the library also supports Appmethod.
Head over and check out the full feature set of the Genetic Algorithms & Programming Component Library for Delphi XE8 Firemonkey.