Markus F. Brameier, Wolfgang Banzhaf | ||||
Springer | ||||
2007 | ||||
English | ||||
315 pages | ||||
1.74 MB | ||||
[tab] [content title="Summary"] Linear Genetic Programming (LGP) focuses on the evolution of imperative computer programs represented as linear sequences of instructions. Unlike traditional Genetic Programming (GP), which utilizes functional expressions or syntax trees, LGP uses a linear program structure as genetic material. This approach leverages its primary characteristics to enhance both execution speed and evolutionary advancement. By conducting online analysis and optimization of program code, LGP develops more efficient techniques and deepens the understanding of the methodology and its parameters. Notably, reducing the step size of structural variation and minimizing non-effective variations are crucial for discovering higher-quality, less complex solutions. This volume explores typical GP phenomena, such as non-effective code, neutral variations, and code growth, specifically through the lens of linear GP. The text is structured into three sections, each detailing methodologies and showcasing applications. Part I introduces fundamental concepts of linear GP and presents effective algorithms for analyzing and optimizing linear genetic programs during runtime. Part II delves into the design of efficient LGP methods and genetic operators inspired by findings from Part I. Part III examines more advanced techniques and phenomena, including effective step size control, diversity management, code growth, and neutral variations. This book serves as a solid introduction to linear GP while offering a comprehensive exploration of its principles and techniques. It will undoubtedly be an invaluable resource for both researchers and students in the field. [/content] [content title="Content"] [/content] [content title="Author(s)"] [/content] [/tab]
[facebook src="bibliosciencesorg"/]
Key-Words: Télécharger Linear Genetic Programming EBOOK PDF EPUB DJVU . Download EBOOK PDF EPUB Linear Genetic Programming DJVU .