Genetically breeding populations of computer programs to solve problems in artificial intelligence

Koza, John R.

doi:10.1109/tai.1990.130444

Cited by 88 publications

(83 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Therefore, the genetic algorithms methodology (GA) has been selected for the optimization by the reason that GA is very useful in this kind of problems where the functions are difficult or impossible to differentiate and cannot be easily solved by analytic procedure (Holland 1975;Koza 1990).…”

Section: Genetic Algorithms Methodologymentioning

confidence: 99%

Development of a hinge compatible with the kinematics of the knee joint

Bertomeu¹,

Lois

Guillem

et al. 2007

Prosthetics &Amp; Orthotics International

View full text Add to dashboard Cite

This study aims to present a new concept of a knee hinge based on a crossed four-bar linkage mechanism which has been designed to optimally follow a motion curve representing the knee kinematics in the position at which the knee hinge should be placed. The methodology used to determine the optimal knee hinge is based on the optimization of certain variables of the crossed four-bar mechanism using genetic algorithms in order to follow a certain motion curve, which was determined using a biomechanical model of the knee motion. Two current, commercially available knee hinges have been used to theoretically determine their motion by means of the path performed by their instantaneous helical axis. Comparison between these two different knee hinges, Optimal Knee Hinge and the theoretical motion performed by a human knee reveals that a common monocentric hinge has a maximum misalignment of up to 27.2 mm; a polycentric hinge has a maximum misalignment of 23.9 mm. In contrast, the maximum misalignment produced by the Optimal Knee Hinge is 1.99 mm. The orthotic joint presented significantly improves the kinematical compatibility and the adjustment between orthotic and human joint motion, and should provide several advantages in terms of comfort and safety. Furthermore, the determination of the instantaneous helical axis for a particular user, by means of human movement measurement techniques, will enable the optimal crossed four-bar mechanisms to be determined in a customized and personalized manner. As a consequence, this new concept of orthotic knee joint design may improve the adaptability of lower limb orthoses for the user, and may lead to significant advantages in the field of orthotics for the lower limb.

show abstract

Section: Genetic Algorithms Methodologymentioning

confidence: 99%

Development of a hinge compatible with the kinematics of the knee joint

Bertomeu¹,

Lois

Guillem

et al. 2007

Prosthetics &Amp; Orthotics International

View full text Add to dashboard Cite

show abstract

“…Genetic Programming (GP) [8] is an evolutionary algorithm in which the evolutionary units are computer programs or functions described by tree structures consisting of conditional branches, mathematical operators, variables and constants [2]. We based GenProgAgent on the generational genetic algorithm [6].…”

Section: Genprogagent: Genetic Programmingmentioning

confidence: 99%

Comparing Apples with Oranges: Evaluating Twelve Paradigms of Agency

Luotsinen

Ekblad

Fitzgibbon

et al. 2007

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. We report on a study in which twelve different paradigms were used to implement agents acting in an environment which borrows elements from artificial life and multi-player strategy games. In choosing the paradigms we strived to maintain a balance between high level, logic based approaches to low level, physics oriented models; between imperative programming, declarative approaches and "learning from basics" as well as between anthropomorphic or biologically inspired models on one hand and pragmatic, performance oriented approaches on the other. Instead of strictly numerical comparisons (which can be applied to certain pairs of paradigms, but might be meaningless for others), we had chosen to view each paradigm as a methodology, and compare the design, development and debugging process of implementing the agents in the given paradigm. We found that software engineering techniques could be easily applied to some approaches, while they appeared basically meaningless for other ones. The performance of some agents were easy to predict from the start of the development, for other ones, impossible. The effort required to achieve certain functionality varied widely between the different paradigms. Although far from providing a definitive verdict on the benefits of the different paradigms, our study provided a good insight into what type of conceptual, technical or organizational problems would a development team face depending on their choice of agent paradigm.

show abstract

“…Genetic Programming, an ensemble of techniques originally developed by Koza,12 applies the Darwinian natural selection paradigm to the evolution of computer programs or symbolic expressions. It is a powerful method for transforming a high level problem statement into a sequence of instructions that can be run on a computer.…”

Section: Evolution Of Intervening Variables Using Genetic Programmentioning

confidence: 99%

Evolving Intervening Variables for Response Surface Approximations

Sóbester¹,

Nair²,

Keane³

2004

10th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference

View full text Add to dashboard Cite

Genetic Programming (GP) is a powerful string processing technique based on the Darwinian paradigm of natural selection. Although initially conceived with the more general aim of automatically producing computer code for complex tasks, it can also be used to evolve symbolic expressions, provided that we have a fitness criterion that measures the quality of an expression. In this paper we present a GP approach for generating functions in closed analytic form that map the input space of a complex function approximation problem into one where the output is more amenable to linear regression. In other words, intervening variables are evolved in each dimension, such that the final approximation model has good generalization properties and at the same time, due to its linearity, can easily be incorporated into further calculations. We employ least squares and cross-validation error measures to derive the fitness function that drives the evolutionary process. Results are presented for a one-dimensional test problem to illustrate some of the proposed ideas -this is followed by a more thorough empirical study, including multi-dimensional approximations and an engineering design problem.

show abstract

Genetically breeding populations of computer programs to solve problems in artificial intelligence

Cited by 88 publications

References 3 publications

Development of a hinge compatible with the kinematics of the knee joint

Development of a hinge compatible with the kinematics of the knee joint

Comparing Apples with Oranges: Evaluating Twelve Paradigms of Agency

Evolving Intervening Variables for Response Surface Approximations

Contact Info

Product

Resources

About