Population-based algorithms using estimation of distribution, often called estimation of distribution algorithms (EDAs), have been recognized as a major paradigm in evolutionary computation. Some EDA-like algorithms have achieved state-of-the-art performance in applications. However, most of the existing EDA-like algorithms have been developed based on an ad-hoc basis. Both theory and implementation in this area are far from complete. This project will work towards establishing a sound theory for characterizing and explaining EDA-like algorithms. We will extend our previous results and seek to obtain global convergence conditions for the factorized distribution algorithm and ant-colony optimization algorithms. With a better understanding of the working mechanisms of EDA-like algorithms, we will embed experimental design methods in EDA to develop more powerful and statistically sound search and optimization algorithms. We will also exploit the potential benefits of combining EDA-like algorithms with GLS (Guided Local Search). We will propose to use continuous optimization problems and quadratic assignment problems as our test problems.
This project is funded by EPSRC (GR/R64742) and University of Essex.
A good introduction of EDA can be found in Professor Pedro Larraņaga's group at University of the Basque Country, Spain, or via links in Qingfu Zhang's page
Updated: 28 December 2003