GEP Book

  Home
  News
  Author
  Q&A
  Tutorials
  Downloads
  GEP Biblio
  Contacts

  Visit Gepsoft

 

C. FERREIRA In H. J. Caulfield, S.-H. Chen, H.-D. Cheng, R. Duro, V. Honavar, E. E. Kerre, M. Lu, M. G. Romay, T. K. Shih, D. Ventura, P. P. Wang, Y. Yang, eds., Proceedings of the 6th Joint Conference on Information Sciences, 4th International Workshop on Frontiers in Evolutionary Algorithms, pages 614-617, Research Triangle Park, North Carolina, USA, 2002.

Mutation, Transposition, and Recombination: An Analysis of the Evolutionary Dynamics

Transposition
 

As shown in Figure 1 and, against all expectations, RIS transposition is more efficient than IS transposition. Recall that with RIS transposition the root itself is always the target, modifying drastically the expression trees [3]. Note also that the transforming power of this kind of transposition is slightly less than mutation but superior to crossover.

Figure 3 compares the evolutionary dynamics obtained for small and high transposition rates. Note the appearance of the type Healthy And Strong for pris = 1.0 (plot b) and for pis = 1.0 (plot d). The other plots were obtained for pris = 0.1 (plot a) and pis = 0.1 (plot c), and are of the kind Healthy But Weak. In fact, and as observed for mutation, an increase in transposition rate results in an increase in the gap between best and average fitness.

Figure 3. A gallery of evolutionary dynamics found in populations evolving by RIS (plots a and b) and IS transposition (plots c and d). The success rate above each plot was determined in the experiment shown in Figure 1. a) Healthy But Weak dynamics, pris = 0.1. b) Healthy And Strong dynamics, pris = 1.0. c) Healthy But Weak dynamics, pis = 0.1. d) Healthy And Strong dynamics, pis = 1.0.

Home | Contents | Previous | Next