Genetic Algorithms

Definition:

Models that optimize rules by mimicking the Darwinian Law of survival of the fittest. A set of rules is chosen from those that work the best. The weakest are discarded. In addition, two successful rules can be combined (the equivalent to genetic cross-overs) to produce offspring rules. The offspring can replace the parents, or they will be discarded if less successful than the parents. Mutation is also accomplished by randomly changing elements. Mutation and cross-over occur with low probability, as in nature.

Investing Essentials


Copyright © 2011 Campbell R. Harvey, Professor of Finance, Fuqua School of Business at Duke University

Term of the Day

Portfolio opportunity set

The expected return/standard deviation pairs of all portfolios that can be constructed from a given set of assets.

Subscribe to the Term of the Day via email Get the Term of the Day in your inbox!


Create your free portfolio