报告人:日本理化学研究院(RIKEN)先进智能研究中心赵启斌研究员
报告题目:New progress in Tensor based machine learning: algorithms and applications
报告时间:2018年19月15日(周六)下午15:30-16:30
报告地点:一教500
欢迎广**生参会学习交流!
报告摘要:
Abstract:Tensor (multiway-array) enables us to represent high-dimensional data efficiently and naturally. Tensor decompositions and tensor networks are emerging technologies in machine learning field in recent years. Tensor models are usually powerful for representing high-dimensional structured data, or high order relations, improving the computational efficiency, handling large-scale optimization problems.In this talk, I will present firstly the basic tensor algebra and decomposition models, then the various algorithms are introduced as well as the real-world applications. Finally, I will present some ongoing projects related to tensor networks and deep learning.
报告人简介:
赵启斌教授2009年获得上海交通大学博士学位,2009年赴日本理化学研究院(RIKEN)脑科学所从事脑信号处理方面的研究,2016年加入刚成立的日本理化学研究院(RIKEN)先进智能研究中心任研究室主任,从事人工智能理论及应用研究。赵博士主要从事领域包括机器学习、张量分析、神经计算、脑-机接口及计算机视觉等,已发表学术论文30余篇,在顶级期刊TPAMI发表论文两篇。