Genetic Programming Theory and Practice XII (Genetic and Evolutionary Computation)
Format: PDF / Kindle (mobi) / ePub
These contributions, written by the foremost international researchers and practitioners of Genetic Programming (GP), explore the synergy between theoretical and empirical results on real-world problems, producing a comprehensive view of the state of the art in GP. Topics in this volume include: gene expression regulation, novel genetic models for glaucoma, inheritable epigenetics, combinators in genetic programming, sequential symbolic regression, system dynamics, sliding window symbolic regression, large feature problems, alignment in the error space, HUMIE winners, Boolean multiplexer function, and highly distributed genetic programming systems. Application areas include chemical process control, circuit design, financial data mining and bioinformatics. Readers will discover large-scale, real-world applications of GP to a variety of problem domains via in-depth presentations of the latest and most significant results.
selection proceeds in two stages modeled after Pareto domination tournaments and fitness sharing described by Horn et al. (1994). As described in detail (Moore et al. 2013), we used classification accuracy, the number of features or attributes in the model (i.e. complexity) and interaction information (Moore et al. 2006) as the axes in the Pareto optimization. Here, the sum of the interaction information for all pairs of attributes in a model is the measure of interestingness (described in more
types, for which we would have to hand-code the map functions. In languages with pattern-matching primitives, such as Charity (see home page at http://pll.cpsc.ucalgary.ca/charity1/www/home.html), you effectively get fold and map functions for free. Whether this can be achieved in a pure combinator language remains to be seen. Note that Y cannot be assigned a valid type, since otherwise there would be a contradiction to the strong normalization theorem. Another approach is to introduce a
Gerstein 2012). With these models, we aim to investigate how much variation of gene expression levels can be explained by TF binding and histone modification signals, respectively. We have tested the models in multiple species ranging from yeast to human. Here we show the results using data obtained from mouse embryo stem cells (mESCs). Specifically, the data contain ChIP-seq profiles for 12 TFs (E2f1, Esrrb, Klf4, Nanog, Oct4, Stat3, Smad1, Sox2, Tcfcp2l1, Zfx, c-Myc and n-Myc) and 7 histone
Helmuth T (2011b) Tag-based modules in genetic programming. In: GECCO ’11: Proceedings of the 13th annual conference on genetic and evolutionary computation, ACM, Dublin, Ireland, pp 1419–1426 Spector L, Harrington K, Helmuth T (2012) Tag-based modularity in tree-based genetic programming. In: GECCO ’12: Proceedings of the 14th international conference on Genetic and evolutionary computation conference, ACM, Philadelphia, Pennsylvania, USA, pp 815–822 Uy NQ, Hoai NX, ONeill M, McKay RI,
does the regression sees — all data points are perfectly accurate. In our analysis of the HUMIES (Tables 9.1, 9.2, 9.3, 9.4, 9.5), we explicitly chose to ignore whether the entry won a gold, silver, or bronze award, since we believe that this is insignificant to the analysis because an entry that has won any award at all is necessarily human-competitive. 9.4 Lessons Learned First and foremost, we note that techniques from evolutionary computation have been used to solve problems from a very