ALIFE conference to reveal artificial agents that learn better
A new way to regulate the extent to which artificial agents rely on social learning will be discussed at the first European conference on Artificial Life, hosted by ECS next month.
Dieter Vanderelst, Rene Ahn and Emilia Barakova from Eindhoven University of Technology will present a paper entitled Simulated Trust: Towards robust social learning on Tuesday 5 August. They will describe how they have developed a mechanism which makes it possible for artificial agents to regulate their reliance on social learning.
According to the researchers, although social learning is a potentially powerful learning mechanism to use in artificial multi-agent systems, findings in the animal kingdom show that it is also possibly detrimental as it could lead to agents acting on second-hand information that might not be trustworthy.
The researchers' simulations have shown that this new proposed trust mechanism is effective in regulating the extent to which agents rely on social learning and causes considerable improvements in their learning rate. (A copy of the paper is available from Joyce Lewis at: email@example.com).
The newly-formed Science and Engineering of Natural Systems (SENSe) group within the University of Southampton's School of Electronics and Computer Science (ECS) is hosting this year's conference, which will take place at the University of Winchester West Downs Campus, involving 250 participants and more paper presentations than ever before.
"This is a critical time for Artificial Life," said Dr Seth Bullock at ECS, the conference chair. "The field is on the verge of synthesising living cells, a feat that the Artificial Life community could only dream of when it started out in the late 80s."
Keynote speakers include internationally leading experts such as Professor Stuart Kauffman, author of The Origins of Order, Professor Peter Schuster, editor-in-chief of the journal Complexity, Professor Eva Jablonka, author of Evolution in Four Dimensions (with Marion Lamb), and Professor Andrew Ellington, a leading pioneer in the new science of synthetic biology.
Professor Takashi Ikegami from the University of Tokyo will open the conference, speaking on work spanning self-organisation and autopoiesis in systems of birds, robots, children, flies, cells, and even oil droplets. The conference is unified by a focus on understanding the fundamental behavioural dynamics of embedded, embodied, evolving and adaptive systems.