报告人:黄鹤
报告地点:腾讯会议ID:889 864 8458
报告时间:2022年10月18日星期二14:00-15:00
邀请人:徐东坡
报告摘要:
In this talk, we will firstly introduce some preliminaries on feedforward complex-valued neural networks. Then, two kinds of adaptive learning algorithms for feedforward complex-valued neural networks are presented, including adaptive complex-valued stepsize based gradient descent algorithm and adaptive complex-valued L-BFGS algorithm. The efficiency of the two learning algorithms are verified by some experimental results on function approximation, pattern classification and nonlinear channel equalization, etc.
会议密码:115119
主讲人简介:
黄鹤, 苏州大学教授,多次受邀到香港城市大学和德州农机大学卡塔尔分校进行合作研究。主持完成国家自然科学基金项目和江苏省自然科学基金面上项目各2项,入选2016年江苏省高校“青蓝工程”优秀青年骨干教师培养对象,在科学出版社出版专著1部,在国际学术期刊如IEEE Transactions on Neural Networks and Learning Systems, IEEE Transactions on Cybernetics, Neural Networks及国内外学术会议上发表学术论文近百篇,授权发明专利10余项。目前担任中国电子学会主办的《新一代信息技术》的编委,以及国际学术期刊Neurocomputing、Circuits, Systems & Signal Processing和Neural Processing Letters的编委或副主编。