In-house expertise

The initial guidelines for the Centre defined its focus on ‘Cognitive Materials’. Several material systems and areas of expertise were already being studied within the Zernike Institute. The researchers working on these topics are therefore central participants, with leading roles in current research. The materials science contributions aim to explore, study and further design (opto-)electronic materials that can both transfer signals over short timescales and exhibit learning effects over long timescales (e.g. as is present in established memristor functionality). Thereby, the material is suited to the cognitive processing of information, with learning features (adaptivity/plasticity). The long-term learning dynamics may come from effects such as ion displacement and phase changes, while short-term signal transfer typically concerns electronic or optical transport properties. Operating modes with low-energy consumption and toggling (gradual) between learning modes and operational modes, may occur via the tuning of the spiked character of signals or via the gating of field-effects in the material.

We aim to explore this functionality in various material systems and devices: i) where both aspects are intrinsically present at the nanoscale within a material; ii) in hybrid systems where part of the functionality is in the material, while additional transistors are present for feedback routes and tuning; iii) through a fully device-based approach, particularly relevant to devices that have the prospect of dense integration and very low power consumption. In this regard, we are building on our expertise in materials with tuneable conducting domain walls, skyrmions, functionalized carbon nanotubes, phase change or ferroic materials, as well as expertise on optoelectronics, spintronics, thermoelectrics, polymer self-assembly, nano-ionics and device physics.

Prof. dr. Tamalika Banerjee
Spintronics of Functional Materials

Prof. dr. Elisabetta Chicca
Bio-Inspired Circuits and Systems

Dr. Erika Covi
Cognitive Devices

Prof. dr. ir. Bart Kooi
Nanostructured Materials and Interfaces

Prof. dr. Maria Loi
Photophysics and OptoElectronics

Prof. dr. Beatriz Noheda
Nanostructures of Functional Oxides

Prof. dr. George Palasantzas
Physics (Surface interactions and Nanostructures)

Prof. dr. Petra Rudolf
Experimental Solid State Physics

The mathematics expertise present at the University of Groningen covers a broad spectrum (Statistics & Stochastics, Systems & Control, Computational Mathematics and Dynamical Systems). Mathematics is already closely connected to the computer sciences, for example at the systems and engineering level, through Systems & Control. Other contributions to the Centre involve the modelling of the large-scale behaviour of stochastic systems; control analysis of large-scale complex systems; and large-scale simulations and numerical analysis.

An overarching theme of the mathematics department is the analysis and control of dynamic systems. These systems can be autonomous or open to interaction with other systems. The subject involves a variety of mathematical theories ranging from analysis and algebra to geometry and measure theory. Statistical, stochastic and algebraic aspects of network dynamics also play an important role. Dynamic systems theory and systems and control theory are used throughout the natural and engineering sciences, from mathematical physics to the earth and life sciences, and from fluid dynamics to power networks. Another important theme across the department is computational and algorithmic mathematics, linking mathematical and physical modelling, the simulation of dynamics, geometric computing and analysing networks.

Dynamic systems theory is concerned with the behaviour of systems that evolve over time. Above all, this concerns the long-term behaviour that comprises stationary, periodic, multi-periodic and chaotic dynamics, but transient behaviour is also of interest. Moreover, bifurcations or transitions between asymptotic states – in particular transitions between regular and chaotic motions – under the variation of parameters are of great importance. Mathematics in Groningen is developing mathematical tools using methods from analysis, geometry and measure theory to grasp, study and develop the structures involved. Moreover, it is developing methods to detect and understand the dynamics in specific models, employing numerical and visualization tools and computer algebra.

Algorithmic and constructive methods also play an important role. If the number of degrees of freedom is huge, then such systems are best described by statistical means. Statistical mechanics deals with the question of how global observables, such as temperature, can be explained based on microscopic behaviour. There is a close relationship with dynamic systems theory, in particular with regard to random and chaotic behaviour and what are called ‘non-equilibrium systems’. When viewing such systems from an experimental point of view, the old paradigm of many observations relative to the number of predictors is obsolete. New high-dimensional inference methods based on sparsity and borrowing strength have become essential to face such challenges. The analysis, control and optimization of such complex systems are studied in the highly esteemed Systems, Control & Optimization group. There are significant opportunities for close collaboration between materials engineering and mathematics in Groningen, both through fundamental mathematical research and in collaboration with colleagues from engineering and the natural sciences.

Prof. dr. Bart Besselink
Systems and Control Theory

dr. Gilles Bonnet
Stochastics

Dr. Serte Donderwinkel
Probability Theory

Dr. Julian Koellermeier
Computational Mathematics

Prof. dr. Marco Grzegorczyk
Computational Statistics

Prof. dr. Arjan van der Schaft
Applied Analysis

Dr. Alef Sterk
Dynamical Systems Theory

Prof. dr. Holger Waalkens
Dynamical Systems Theory

Prof. dr. ir. Fred Wubs
Numerical Mathematics

In terms of its existing expertise, Computer Science at the University of Groningen is well positioned to make a significant contribution to the Centre. There are several groups working in Image Processing & Computer Vision that are developing state-of-the-art morphological image operators for feature extraction and description from very large images and image sequences. This expertise can be harnessed to develop efficient image analysis algorithms to analyse conductivity levels and conduction paths in images of nanomaterials.

Computer Science has a long tradition in developing biologically motivated and brain-inspired pattern recognition and machine-learning methods, which is directly relevant to Cognitive Computing. Expertise in Computer Graphics, Visualization and Visual Analytics can be applied to the visualization of computational infrastructure (system, pipelines and networks), processes and data in order to support the interactive design of cognitive materials and gain insight into the complex processes and structures involved. Systems engineering expertise is available for the design of very complex, scalable and/or distributed systems-of-systems, such as cognitive systems that comprise heterogeneous and operationally independent constituent systems.

In Fundamental Computing, the expertise covers areas such as logic, discrete structures, advanced algorithms and data structures, and the formal modelling of communicating systems. This knowledge is needed to develop new computing paradigms, algorithms and programs for the new cognitive systems, and to understand the computational complexity of such systems in a precise mathematical sense.

There are also cross-links with complementary expertise in AI in relation to machine learning, automated reasoning and human-computer interaction. Collaboration with the materials scientists has already been established for the design of cognitive materials (efficient image analysis algorithms for analysing conductivity levels and conduction paths during network training). Other direct contributions will involve the use of machine learning and pattern recognition in cognitive system design, interactive network and system visualization and parameter inference in complex systems.

Prof. dr. George Azzopardi
Computer Science

Prof. dr. Michael Biehl
Intelligent Systems

Prof. dr. ir. Georgi Gaydadjiev
Innovative Computer Architecture

dr. Revantha Ramanayake
Theory of Computation

Prof. dr. Jos Roerdink
Scientific Visualization and Computer Graphics

Dr. Fatih Turkmen
Computer and Network Security

Dr. Michael Wilkinson
Digital image analysis and computer vision

The University of Groningen holds a strong position in artificial intelligence (knowledge-based and multi-agent systems), pattern recognition/machine learning and cognitive modelling/data-science engineering. We have expertise in neural networks, both in artificial and natural systems, in cognitive neuro-scientific modelling and computational neuroscience. Problems in cognitive systems range from the materials of the computing substrate to high-level cognitive mechanisms and applications.

In recent years, the availability of large data sets and computing power have led to a revolution in machine learning, notably in the area of ‘Deep Learning’ with neural networks (Bengio, 2009; Bengio et al., 2015). Improvements in learning algorithms now allow for neural networks with more layers and parameters (weights) than ever before. For patterns in 2D or 3D arrays, the use of convolution kernels allows a saving in the number of coefficients and a gain in terms of robustness against pattern variation. The concept of a ‘universal function approximator’ has been mathematically supported (Cybenko, 1989; Hornik, 1999).

This state of the art leads to opportunities and challenges. Deep learning now allows the detection, classification and prediction of spatiotemporal patterns in continuous and discrete data. This allows for a functionality that is now called Cognitive Computing, aimed at uncovering meaningful knowledge from raw data. However, there are also several challenges. First, it should be noted that current neural-networking methods are implemented on Turing/Von Neuman machines with a very high-energy demand, both for computing and communication functions. At the same time, the human brain, with 9x10¹⁰ neurons (1.5x10¹⁴ connections), uses less energy than a light bulb. Current GPUs for neural-network training may be 1000 Watts each, and their information-processing capacity is still limited in comparison to the human brain. Notably, a snapshot of electrical brain activity will reveal that most neurons are silent. The neuron can be considered a spike oscillator (point-process generator), firing only when necessary.

To discover and exploit new materials with nonlinearity and adaptive connectivity, it is necessary to determine the fundamental classes of computing that are suitable candidates for neuromorphic computing, using both mathematical theory and simulation. Finally, despite the success of deep learning, there are theoretical and practical challenges: How do we prevent and detect inappropriate instances of learned models? How do we learn from raw data with a minimum of training examples? In close cooperation with the materials scientists, we will develop models and methods facilitating the search for materials that are suitable for neural computing due to their electrical nonlinearity and ability for trace formation. In this way, we can cover the full range from low-level modelling to cognitive principles. Our strong international connections with researchers at the CWI (Centrum voor Wiskunde en Informatica) in Amsterdam (Sander Bohte) and at the University of Waterloo (Chris Eliasmith) are highly beneficial in this endeavour.

Dr. Matthew Cook
Cortex-inspired Computing

Prof. dr. Davide Grossi
Cognitive Multiagent Systems

Prof. dr. Herbert Jaeger
Computation in Cognitive Materials

Prof. dr. Lambert Schomaker
Artificial Intelligence

Prof. dr. Niels Taatgen
Artificial Intelligence

Prof. dr. Bart Verheij
Computational argumentation

Marieke van Vugt
Cognitive Modeling