Seminar “Artificial Intelligence and Big Data: the role of machine learning”

Thời gian: 14.30 ngày thứ 2, 6/11/2017
Địa điểm: phòng 803, nhà B1, Đại học Bách khoa Hà Nội
Người trình bày: GS. Geoffrey Holmes, University of Waikato, New Zealand

Title: Artificial Intelligence and Big Data: the role of machine learning

Abstract: The two terms AI and Big Data are now used freely by journalists to typically describe advances in robotics using vision systems and drive fears of big brother – Google knows more about you than your partner, for example. I argue that the big advances come from machine learning and chart the development of the main contributing ideas and methods. I then introduce our own work on application development and big data processing which has centered around the production of open-source software. I conclude with some thoughts on where the field is heading and where more development effort is required.

Speaker:

Prof Geoffrey Holmes is the Dean of Computing & Mathematical Sciences, University of Waikato, New Zealand. Professor Holmes received his degrees in Mathematics from the University of Southampton, England. His PhD involved the development of software packages to assist Mathematicians in solving Einstein’s field equations in General Relativity. This was how he got started in Computer Science. After graduating he became a Research Assistant at the Electrical Engineering Department of Cambridge University, England where he was a member of large team working on a speech understanding system. He took up a position as Lecturer in Computer Science in 1987. In 2008 he was appointed Dean Computing & Mathematical Sciences.

Prof. Geoff Holmes has, in the past, been head of the machine learning group at the University of Waikato and has been involved in several open source projects over the last 20 years. Waikato’s machine learning project has had a far-reaching influence on developments in the field worldwide, principally through the open-source Weka software, one of the most widely used machine learning tools in the world today (Weka software has been downloaded 6 million times since it was first hosted at the Sourceforge website for open-source software in April 2000 – currently at a rate of about 3,000 downloads a day). Academic publications (close to 400 articles) by the machine learning group can be found at http://www.cs.waikato.ac.nz/ml/publications.html.

Professor Holmes has led the applied machine learning subgroup at the University for the past 10 years. This group has particular expertise in the deployment of machine learning solutions in practice and has developed a bespoke platform for this purpose (see https://adams.cms.waikato.ac.nz/).

Professor Holmes has also been a co-developer of a platform for processing very large (possibly infinite) datasets MOA (Massive Online Analysis), which is to data stream mining what Weka is to batch learning (see http://moa.cms.waikato.ac.nz/).

Aside from software, Professor Holmes has made contributions to the major conferences in machine learning and data mining (over 120 academic publications). He has recently co-written a book describing the techniques contained in the MOA software. He was part of the team that in 2005 won the SIGKDD Data Mining and Knowledge Discovery Service Award for Weka and regularly serves on senior program committees for KDD, ECMLPKDD and Discovery Science and referees articles in all the major journals in the field.

Leave a Reply

Your email address will not be published. Required fields are marked *