Genetic programming

Autori

• Pierre Collet

Parametri

• 361 pagine
• 13 ore di lettura

• Computer, IT, programmazione

Maggiori informazioni sul libro

The content explores various applications and advancements in genetic programming (GP) and evolutionary algorithms. It includes detection of 802.11 de-authentication attacks, strategies to enhance efficiency and success rates in GP, and approaches to Solomonoff’s probabilistic induction. The text discusses a context-aware crossover operator, AQUAGP for approximate query answers, and the Blindbuilder encoding for evolving structures. It covers dynamic scheduling, emergent locomotion gaits in simulated robots, and the evolution of crossover operators for function optimization. The role of validation sets and parsimony pressure in GP is examined, along with geometric crossover for biological sequences and methods to manage constraints in evolutionary algorithms. Additionally, it highlights iterative filter generation, prototype optimization, and learning recursive functions with object-oriented GP. The concept of negative slope coefficients to characterize fitness landscapes and population clustering is also addressed. Applications in financial data projection, Sudoku solving, and the halting probability in Von Neumann architectures are discussed. The dynamics of GP are investigated through subtree crossover distance, while diversity characterization and complexity in Cartesian GP are analyzed. Lastly, it touches on robust communication systems design, developmental evaluation in GP, and optimizing decision heuristics in SAT sol

Lingua: Tedesco

Pubblicato: 2006

(In brossura)