Information Theory in Artificial Intelligence and Artificial Life
We study principled methods (with focus on information theory) in fundamental questions of Artificial
Intelligence and Artificial Life, including principles of cognition, and emergence (origin) of life and intelligence from first principles. Questions also include information processing in adaptive, complex and
self-organizing systems in general, a research area which has witnessed a dramatic increase of interest
in the last years. At the same time, our goal is to adapt these principles for AI applications and robotic
devices.
We use mathematical methods, with particular emphasis on Shannon’s information theory to describe,
understand or construct systems in the context of AI/robotics and biology. Potential research directions
could be:
• information and its flow through the perception-action loop
• the role of information in the self-organization in complex systems
• origin of life and cognition
• principled Artificial Intelligence methods for embodied autonomous robotics
• intrinsic motivation and artificial creativity
• informational principles of biological computation (with opportunities to collaborate with the Biocomputation Research Group)
The prospective candidates should have a very strong first degree, a keen interest in delving into and
contributing to a new, and quickly expanding research area and an outstanding background in Computer
Science, Physics, Mathematics, Statistics or another relevant computational/quantitative discipline. In
particular, they should demonstrate excellent programming skills in at least one major computer language.
A mathematical/numerical background would be highly desirable. Knowledge in at least one of the
following fields would be a plus: probability theory, information theory, differential geometry, control,
dynamical systems, control, data modelling/neural network techniques.
The envisaged research will take place in the vibrant, enthusiastic and enterprising research environment of the Adaptive Systems Research Group in the School of Computer Science at the University of
Hertfordshire; there will also be the chance to collaborate with the socSMCs (Socializing Sensorimotor
Contingencies, FET Open) and WiMUST (Widely Scalable Mobile Underwater Sonar Technology) EU
Horizon 2020 projects, as well as the School’s successful humanoid robot RoboCup team, the Bold Hearts
and other projects at the School.
Contact:
Prof. Dr. Daniel Polani
[email protected]