-
6月29日学术报告2(华东师范大学 周爱民:Learning Guided Evolutionary Multiobje
-
类别：网络整理
发布人：admin
浏览次数： 次
发布时间：2015-09-02 14:24
-
报告题目： Learning Guided Evolutionary Multiobjective Optimization
报告日期及时间：2015年6月29号上午10：00
报告地点： B403
报告人：周爱民
报告人单位：华东师范大学
报告人简介：
周爱民，华东师范大学副教授。分别于2001年和2003年在武汉大学获得计算机学士和硕士学位，2009年在英国Essex大学获得计算机博士学位，2009年起在华东师范大学工作。目前发表40余篇学术论文，包括IEEE Transactions on Evolutionary Computation, IEEE Transactions on Cybernetics, GECCO, CEC, EMO等期刊和会议。目前担任期刊Swarm and Evolutionary Computation副主编。主要从事演化算法与优化、机器学习、(遥感)图像处理及应用等领域的研究与教学工作。
报告摘要： The population of an evolutionary algorithm can be regarded as a data set that contains some kind of patterns. Although
some evolutionary algorithms, such as estimation of distribution algorithms which utilize the probability graphic models to extract
the patterns and surrogate assisted evolutionary algorithms which use regression methods, try to find the patterns and to guide
the evolution process, there is still lack of a systematic work on using machine and statistical learning techniques to guide the
evolutionary optimization. Furthermore, the area of machine and statistical learning contains a large broad of techniques but only
a few ones have been utilized in the community of evolutionary computation. This talk tries to build a bridge from machine and
statistical learning to evolutionary optimization especially evolutionary multiobjective optimization. The talk will cover the background
on multiobjective optimization, an example on how to use self-organizing maps to assist the search, a short survey on our recent work on this topic, and some conclusions and remarks for future work.
邀请人： 王峰 副教授
-
- 上一篇：8月29日学术报告（国家自然科学基金委员会副主任高文院士：多媒体大数据处理及其技术挑战）
- 下一篇：6月29日学术报告1(北京邮电大学 石川:演化多目标机器学习)
