DROPS

Document

N-gram GP: Early results and half-baked ideas

Authors: Nicholas Freitag McPhee and Riccardo Poli

Published in: Dagstuhl Seminar Proceedings, Volume 8051, Theory of Evolutionary Algorithms (2008)

Abstract

In this talk I present N-gram GP, a system for evolving linear GP programs using an EDA style system to update the probabilities of different 3-grams (triplets) of instructions. I then pick apart some of the evolved programs in an effort to better understand the properties of this approach and identify ways that it might be extended. Doing so reveals that there are frequently cases where the system needs two triples of the form ABC and ABD to solve the problem, but can only choose between them probabilistically in the EDA phase. I present the entirely untested idea of creating a new pseudo-instruction that is a duplicate of a key instruction. This could potentially allow the system to learn, for example, that AB is always followed by C, while AB' is always followed by D.

Cite as

Nicholas Freitag McPhee and Riccardo Poli. N-gram GP: Early results and half-baked ideas. In Theory of Evolutionary Algorithms. Dagstuhl Seminar Proceedings, Volume 8051, pp. 1-3, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2008)

Copy BibTex To Clipboard

@InProceedings{mcphee_et_al:DagSemProc.08051.5,
  author =	{McPhee, Nicholas Freitag and Poli, Riccardo},
  title =	{{N-gram GP: Early results and half-baked ideas}},
  booktitle =	{Theory of Evolutionary Algorithms},
  pages =	{1--3},
  series =	{Dagstuhl Seminar Proceedings (DagSemProc)},
  ISSN =	{1862-4405},
  year =	{2008},
  volume =	{8051},
  editor =	{Dirk V. Arnold and Anne Auger and Jonathan E. Rowe and Carsten Witt},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.08051.5},
  URN =		{urn:nbn:de:0030-drops-14838},
  doi =		{10.4230/DagSemProc.08051.5},
  annote =	{Keywords: Genetic programming, estimation of distribution algorithms, linear GP, machine learning}
}

Document

DOI: 10.4230/DagSemProc.06061.7

On Turing complete T7 and MISC F--4 program fitnes landscapes

Authors: William B. Langdon and Riccardo Poli

Published in: Dagstuhl Seminar Proceedings, Volume 6061, Theory of Evolutionary Algorithms (2006)

Abstract

We use the minimal instruction set F-4 computer to define a minimal Turing complete T7 computer suitable for genetic programming (GP) and amenable to theoretical analysis. Experimental runs and mathematical analysis of the T7, show the fraction of halting programs is drops to zero as bigger programs are run.

Cite as

William B. Langdon and Riccardo Poli. On Turing complete T7 and MISC F--4 program fitnes landscapes. In Theory of Evolutionary Algorithms. Dagstuhl Seminar Proceedings, Volume 6061, pp. 1-28, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2006)

Copy BibTex To Clipboard

@InProceedings{langdon_et_al:DagSemProc.06061.7,
  author =	{Langdon, William B. and Poli, Riccardo},
  title =	{{On Turing complete T7 and MISC F--4 program fitnes landscapes}},
  booktitle =	{Theory of Evolutionary Algorithms},
  pages =	{1--28},
  series =	{Dagstuhl Seminar Proceedings (DagSemProc)},
  ISSN =	{1862-4405},
  year =	{2006},
  volume =	{6061},
  editor =	{Dirk V. Arnold and Thomas Jansen and Michael D. Vose and Jonathan E. Rowe},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.06061.7},
  URN =		{urn:nbn:de:0030-drops-5956},
  doi =		{10.4230/DagSemProc.06061.7},
  annote =	{Keywords: Genetic programming}
}

Search Results

Documents authored by Poli, Riccardo

N-gram GP: Early results and half-baked ideas

Abstract

Cite as

On Turing complete T7 and MISC F--4 program fitnes landscapes

Abstract

Cite as

Thanks for your feedback!

Could not send message