NICE Seminar 7

Surrogate Model-assisted Interactive Evolutionary Computations and Applications

When?

Thursday 17 November 2011, 15:30 to 16:30

Where?

39BB02

Open to:

Public, Staff, Students

Speaker:

Dr Xiaoyan Sun

Interactive evolutionary computation (IEC) is an approach to evolutionary computation (EC) in which user evaluation replaces the fitness function, developed to deal with optimization problems with qualitative indices.  User fatigue resulting from the user’s evaluations, however, restricts the applications of such kinds of algorithms in solving complicated practical problems.  This has recently motivated a lot of efforts to design friendly evaluation styles or build surrogates of the user’s evaluations to effectively alleviate user fatigue.  We have investigated several surrogate-assisted IECs associated with different evaluation modes, and successfully applied them to some practical fields.  In terms of distinct machine learning mechanisms, surrogate models constructed with supervised or semi-supervised learning for different evaluation modes will be focused here.  Some empirical results will be presented to demonstrate the strengths of surrogate-assisted IECs.