Speaker: Dr. Robin Gras - School of Computer Science
Title: An Introduction to Artificial Intelligence and Machine Learning

Abstract :  Artificial Intelligence is a field as old as Computer Science itself. As early as 1950, Alan Turning discussed how to build intelligent machines in his “Computing Machinery and Intelligence” paper. For many years computers have not had the computational power required to run these early models, in particular the connectionist models that would lead to the neural network architecture. Therefore, the progresses in systems that could learn by themselves have been quite slow. It is not before the early 90 that the field of Machine Learning really emerged, through several methods taking their roots from statistical analysis. However, training complex non-linear model such as multi-layers neural networks was still not realistic. Two important improvements occurred in 2006 and 2011 leading to the first successes in efficiently training multi-layers neural networks. The consequences have been huge, with an extremely fast growth in the number, size, and accuracy of these so-called Deep Learning models. It seems today that there is no domain in Artificial Intelligence that can not be tackled by these approaches and new impressive results are published almost every day. This presentation will give an overview of the Machine Learning field, presenting the fundamental concepts and the evolution of the domain up to the very last advances and what are coming next.