A New Method for Generating Nonlinear Correction Models of Dynamic Objects Based on Semantic Genetic Programming

2016 ◽  
pp. 249-261 ◽  
Author(s):  
Łukasz Bartczuk ◽  
Alexander I. Galushkin
Keyword(s):  
Genetic Programming ◽  
New Method ◽  
Nonlinear Correction ◽  
Dynamic Objects

New Method for Nonlinear Fuzzy Correction Modelling of Dynamic Objects

2014 ◽  
pp. 169-180 ◽  
Author(s):  
Łukasz Bartczuk ◽  
Andrzej Przybył ◽  
Petia Koprinkova-Hristova
Keyword(s):  
New Method ◽  
Dynamic Objects

A New Method for Simplifying Algebraic Expressions in Genetic Programming Called Equivalent Decision Simplification

2009 ◽  
Vol 13 (3) ◽  
pp. 237-244 ◽  
Author(s):  
Naoki Mori ◽  
◽  
Bob McKay ◽  
Nguyen Xuan Hoai ◽  
Daryl Essam ◽  
...  
Keyword(s):  
Computer Simulation ◽  
Genetic Programming ◽  
Symbolic Regression ◽  
New Method ◽  
Algebraic Expressions ◽  
Novel Approach ◽  
Key Issues ◽  
Regression Problems

Symbolic Regression is one of the most important applications of Genetic Programming, but suffers from one of the key issues in Genetic Programming, bloat. For a variety of reasons, reliable techniques to remove bloat are highly desirable. This paper introduces a novel approach of removing bloat, Equivalent Decision Simplification, in which subtrees are evaluated over the set of regression points. The effectiveness of the proposed method is confirmed by computer simulation taking simple Symbolic Regression problems as examples.

scholarly journals Empirical Analysis of Schemata in Genetic Programming

2021 ◽  
Author(s):  
◽  
Will Smart
Keyword(s):  
Genetic Programming ◽  
Empirical Analysis ◽  
Global Analysis ◽  
Evolutionary Process ◽  
Building Blocks ◽  
New Method ◽  
Large Set ◽  
Fine Grained ◽  
Analysis System ◽  
The Many

<p>Schemata and buiding blocks have been used in Genetic Programming (GP) in several contexts including subroutines, theoretical analysis and for empirical analysis. Of these three the least explored is empirical analysis. This thesis presents a powerful GP empirical analysis technique for analysis of all schemata of a given form occurring in any program of a given population at scales not previously possible for the kinds of global analysis performed. There are many competing GP forms of schema and, rather than choosing one for analysis, the thesis defines the match-tree meta-form of schema as a general language expressing forms of schema for use by the analysis system. This language can express most forms of schema previously used in tree-based GP. The new method can perform wide-ranging analyses on the prohibitively large set of all schemata in the programs by introducing the concepts of maximal schema, maximal program subset, representative set of schemata, and representative program subset. These structures are used to optimize the analysis, shrinking its complexity to a manageable size without sacrificing the result. Characterization experiments analyze GP populations of up to 501 60- node programs, using 11 forms of schema including rooted-hyperschemata and non-rooted fragments. The new method has close to quadratic complexity on population size, and quartic complexity on program size. Efficacy experiments present example analyses using the new method. The experiments offer interesting insights into the dynamics of GP runs including fine-grained analysis of convergence and the visualization of schemata during a GP evolution. Future work will apply the many possible extensions of this new method to understanding how GP operates, including studies of convergence, building blocks and schema fitness. This method provides a much finer-resolution microscope into the inner workings of GP and will be used to provide accessable visualizations of the evolutionary process.</p>

Sign in / Sign up

Export Citation Format

Share Document