Talk 2

One Day Meeting on Advanced Statistics Topics

CIDMA- University of Aveiro

Speaker: David Hand - Department of Mathematics - Faculty of Natural Sciences, Imperial College London

Title of the talk: Statistics and machine learning: the synergistic intersection

Abstract: The disciplines of machine learning and statistics have very substantial overlap, but with differences in emphasis arising from their origins. Statistics, with roots stretching back hundreds of years, really came into its own in the nineteenth and twentieth century, originally through its role in illuminating societal problems and later as a mathematical discipline when powerful inferential concepts and methods were developed. The development of the computer then facilitated its wider application. In contrast, machine learning, while an ancient dream in contexts such as the sorcerer’s apprentice, really became practicable only with the advent of computers. This led to the development of algorithms alongside what was going on in statistics, in an example of parallel evolution stimulated by cross-fertilization between the two disciplines. As it became apparent that the problems being tackled were often the same, so it also became clear that the two perspectives, one mathematical and one computational, could work in a powerful synergy for progress. Powerful, but ad hoc and intuitively-based computational suggestions found a deep theoretical mathematical underpinning, enabling better understanding and generalisation. I illustrate these developments with examples, showing how statistical ideas lie at the heart of machine learning methods.

SLIDES

David Hand is Emeritus Professor of Mathematics and Senior Research Investigator at Imperial College, London, where he formerly held the Chair in Statistics. He launched the journal Statistics and Computing in 1991, and was one of the founders of the AI and Statistics conference series, which began in 1985 and next year has its 28th meeting. He has published 300 scientific papers and 32 books, including Principles of Data Mining, The Improbability Principle, Dark Data, Information Generation, AI and Computer Power, and Artificial Intelligence Frontiers in Statistics. He has received various awards, including the IEEE-ICDM Research Contributions Award, the George Box medal, and the International Federation of Classification Societies Research Medal.

ORCID: 0000-0002-4649-5622

Page updated

Google Sites

Report abuse