Fri, 04.03.2016
Is it possible to get sense from 'Big Data'?
Prof. Christopher Irgens, PhD
University of Strathclyde, UK
The use of current supervised and unsupervised machine learning methods and tools carry a promise of bringing 'cosmos' to 'chaos'. This seminar is based upon experience from the past 5 - 6 years with 'big data' from two different application domains, namely aerospace manufacturing and financial investment market. In hindsight with varying results.
The methods and associated tools are both supervised and unsupervised, such as Support Vector Machine and Cluster Analysis. These have been used both as single tools and complementary tools with varying results. The theoretical background to these methods and tools is both simple and 'common sense'. This will be 'explored' during both the lecture part and practical part of the seminar. The main focus will be on SVM (Support Vector Machine) and Cluster Analysis. The themes of the seminar will be taken from the above mentioned applications, as will data where possible without breaking confidentiality which is current.
Data-mining will invariably involve considerable effort in the pre-processing of the data, the experience of this will be addressed as will the data volume which brings with it both limitations and opportunities. The theoretical backgrounds of SVM and Cluster Analysis will be introduced and this will be followed by both synthetic and actual cases from completed work.
The practical work will aim at introductory use of available software on prepared data scenarios, so that participants may get some initial understanding of the potential as well as limitations of machine learning.
Short profile
Educated in Norway and England. University of Liverpool from 1966 to 1972 in BEng(HONS) in Mechanical Engineering and MSC in Numerical Analysis and Electronic Computation.
Worked in:
- Avionics in British defence industry
- Early implementations of the POP-2 compiler for Artificial Intelligence use
- Professor and Visiting Professor at British, Norwegian, Swiss, Chzech and Egyptian universities.