skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Genetic algorithms as discovery programs

Abstract

Genetic algorithms are mathematical counterparts to natural selection and gene recombination. As such, they have provided one of the few significant breakthroughs in machine learning. Used with appropriate reward functions and apportionment of credit, they have been successfully applied to gas pipeline operation, x-ray registration and mathematical optimization problems. This paper discusses the basics of genetic algorithms, describes a few successes, and reports on current progress at Oak Ridge National Laboratory in applications to set covering and simulated robots.

Authors:
;
Publication Date:
Research Org.:
Oak Ridge National Lab., TN (USA)
OSTI Identifier:
7058240
Report Number(s):
CONF-8610146-5
ON: DE87002410
DOE Contract Number:
AC05-84OR21400
Resource Type:
Conference
Resource Relation:
Conference: Southeastern TIMS meeting, Myrtle Beach, SC, USA, 9 Oct 1986; Other Information: Paper copy only, copy does not permit microfiche production
Country of Publication:
United States
Language:
English
Subject:
59 BASIC BIOLOGICAL SCIENCES; ALGORITHMS; BIOLOGICAL EVOLUTION; COMPUTERIZED SIMULATION; GENETICS; BIOLOGY; MATHEMATICAL LOGIC; SIMULATION; 550400* - Genetics

Citation Formats

Hilliard, M.R., and Liepins, G. Genetic algorithms as discovery programs. United States: N. p., 1986. Web.
Hilliard, M.R., & Liepins, G. Genetic algorithms as discovery programs. United States.
Hilliard, M.R., and Liepins, G. Wed . "Genetic algorithms as discovery programs". United States. doi:.
@article{osti_7058240,
title = {Genetic algorithms as discovery programs},
author = {Hilliard, M.R. and Liepins, G.},
abstractNote = {Genetic algorithms are mathematical counterparts to natural selection and gene recombination. As such, they have provided one of the few significant breakthroughs in machine learning. Used with appropriate reward functions and apportionment of credit, they have been successfully applied to gas pipeline operation, x-ray registration and mathematical optimization problems. This paper discusses the basics of genetic algorithms, describes a few successes, and reports on current progress at Oak Ridge National Laboratory in applications to set covering and simulated robots.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Wed Jan 01 00:00:00 EST 1986},
month = {Wed Jan 01 00:00:00 EST 1986}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • This paper presents a feature selection methodology that can be applied to datasets containing a mixture of continuous and categorical variables. Using a Genetic Algorithm (GA), this method explores a dataset and selects a small set of features relevant for the prediction of a binary (1/0) response. Binary classification trees and an objective function based on conditional probabilities are used to measure the fitness of a given subset of features. The method is applied to health data in order to find factors useful for the prediction of diabetes. Results show that our algorithm is capable of narrowing down the setmore » of predictors to around 8 factors that can be validated using reputable medical and public health resources.« less
  • This paper discusses the development of a new optimization algorithm for DC motor design. In principle, the new algorithm utilizes a mixed method that consists of genetic algorithms in conjunction with direct search method. The genetic algorithms are used for locating the global optimum region while the direct search method is used to achieve objective function convergence. In order to validate the effectiveness, the new algorithm has been applied to an actual DC motor. Field and torque characteristics of the DC motor are computed using finite element method and the principle of virtual work, respectively.
  • A method to identify the nature of a crack on the surface of a region using nondestructive testing (NDT) and inverse problem methodology is presented. A genetic algorithm (GA) based approach, which involves a global search to avoid local minima, is presented and applied to solve the inverse problem of identifying the position, shape and the orientation of a surface crack. A fine tuning algorithm is combined with the GA to reach the optimum solution.
  • This paper describes a genetic algorithm application to the DNA fragment assembly problems. The genetic algorithm uses a random key representation for representing the orderings of fragments. Two different fitness functions, both based on pairwise overlap strengths between fragments, were tested. The paper concludes that the genetic algorithm is a promising method for fragment assembly problems, achieving usable solutions quickly, but that the current fitness functions are flawed and that other representations might be more appropriate.