当前位置:首页  学术交流

学术讲座【Human Centric Machine Learning -- An overview of previous and current research】

时间:2017-08-11浏览:271设置

时间:2017年8月11日(星期五)上午10:00-11:30

地点:仓山校区成功楼603

主讲:英国阿尔斯特大学 王晖教授

主办:数学与信息学院

专家简介:王晖,教授,英国阿尔斯特大学(Ulster University)人工智能研究中心主任,研究方向包括机器学习和自动推理以及在视频图像光谱分析和决策支持中的应用。近年来主持和参加了20多项欧盟英国北爱资助项目,发表220多篇学术论文,应邀担任20多个国际学术会议的程序委员会委员,3个国际学术期刊的编委。

报告摘要:In this talk I will provide an overview of my previous research in the area of machine learning including lattice machine, contextual probability and neighbourhood counting. Lattice machine is a learning paradigm where the knowledge model is hyper relation which is consistent with human conception of the world according to formal concept analysis. Contextual probability provides a statistical view of any model of this type, which can be used to measure reliability of the model in making predictions. Neighbourhood counting provides means of estimating contextual probability from data. All of these underpin a new point of view, {\em human centric machine learning}, that the knowledge model learned should be human readable and verifiable. Additionally, I will also provide an overview of my current research endeavours.


返回原图
/