inria-00451416, version 1

Bias and variance in continuous EDA

Fabien Teytaud () ^1, 2, 3, Olivier Teytaud () ^2, 3

EA 09 (2009)

• 1:  TAO (INRIA Futurs)
• 
  INRIA – CNRS : UMR8623 – Université Paris XI - Paris Sud France
• 2:  Laboratoire de Recherche en Informatique (LRI)
• http://www.lri.fr/
  CNRS : UMR8623 – Université Paris XI - Paris Sud LRI - Bâtiment 490 Université Paris-Sud 91405 Orsay Cedex France
• 3:  TAO (INRIA Saclay - Ile de France)
• http://tao.lri.fr/tiki-index.php
  INRIA – CNRS : UMR8623 – Université Paris XI - Paris Sud LRI, Bat. 490, Université Paris-Sud, 91405 Orsay Cedex France

Bibliographic reference

• Type of document: Peer-reviewed conferences/proceedings
• Domain: Mathematics/Optimization and Control
• Title: Bias and variance in continuous EDA
• Abstract: Estimation of Distribution Algorithms are based on statistical estimates. We show that when combining classical tools from statistics, namely bias/variance decomposition, reweighting and quasi-randomization, we can strongly improve the convergence rate. All modifications are easy, compliant with most algorithms, and experimentally very efficient in particular in the parallel case (large offsprings).
• Full text language: English
• Publication date: 2009-05-08
• Audience: international
• Conference title: EA 09
• Conference city: Strasbourg
• Country: France
• Conference date: 2009-10-26

Attached file list to this document: 

• inria-00451416, version 1
• http://hal.inria.fr/inria-00451416
• oai:hal.inria.fr:inria-00451416
• From: Fabien Teytaud <>
• Submitted on: Friday, 29 January 2010 09:22:26
• Updated on: Friday, 29 January 2010 22:48:38