Lecturer:Dustin Mixon Fields: Mathematical Data Science
Content
Given several points in a high-dimensional space, we would like to cluster the points according to similarity. But how can one find the best clustering? In this talk, we show how eigenvectors point to the solution.
Lecturer
Dustin Mixon received his PhD in applied and computational mathematics from Princeton University in 2012. He specializes in applied harmonic analysis and mathematical data science.
Lecturer:Terry Stewart Fields: Computational Neuroscience / Machine Learning
Content
It is tricky to get neural networks to efficiently represent values that change over time, or to represent arbitrary spatial shapes and locations. In this talk, we look at two methods for representing time and space, using Legendre Polynomials and Circular Convolution, respectively. We show that they not only have desirable computational properties, but also map well onto Time Cells and Place Cells in the brain.
Dumont, Eliasmith (2020) “Accurate representation for spatial cognition using grid cells” CogSci 2020 http://compneuro.uwaterloo.ca/files/publications/dumont.2020.pdf
Lecturer
Dr. Terry Stewart
Terry Stewart is an Associate Research Officer at National Research Council Canada, working on developing large-scale brain simulations and finding industry applications of such systems using energy-efficient neuromorphic hardware. Previously, Terry was a post-doctoral research associate working with Chris Eliasmith at the Centre for Theoretical Neuroscience at the University of Waterloo. Terry started as an engineer (B.A.Sc. in Systems Design Engineering, University of Waterloo, 1999), did a masters applying experimental psychology on simulated robots (M.Phil. in Computer Science and Artificial Intelligence, University of Sussex, 2000), then a Ph.D. was on cognitive modelling (Ph.D. in Cognitive Science, Carleton University, 2007). Terry is also a co-founder of Applied Brain Research, a research-based start-up company based around using low-power hardware (neuromorphic computer chips) and adaptive neural algorithms.
A network or graph is a way to model objects or organisms and their connections with each other. One example of a network is to connect Facebook users if they are friends on the website. Although the number of Facebook accounts is on the order of 1 billion and the average number of Facebook friends each user has is approximately 100, on average any two users are connected by a chain of less than 4 intermediary Facebook friends. Such a network – where the average number of connections per object is “low” but the average distance between objects is also “low” is called a small-world network. Small-world networks appear in many applications and can influence the spread of disease. In this short talk, we will learn a bit about small-world networks and some examples where they appear.
Literature
“Collective dynamics of ‘small-world’ networks”, Duncan J. Watts & Steven H. Strogatz, Nature volume 393, pages440–442 (1998), https://www.nature.com/articles/30918
“Four Degrees of Separation”, Lars Backstrom, Paolo Boldi, et. al, WebSci ’12: Proceedings of the 4th Annual ACM Web Science Conference, pages 33–42 (June 2012), https://dl.acm.org/doi/10.1145/2380718.2380723
Lecturer
Prof. Emily King
Prof. King (Emily) studies how to best deconstruct things into fundamental building blocks in order to better understand them. This can range from tearing apart neural networks to try to figure out how they work to developing methods for extracting features out of biomedical images to finding building blocks with beautiful mathematical symmetries that are also useful in quantum information. She is abusing the special nature of the 2021 Virtual IK to both give a talk and serve as co-chair, which is usually a big no-no. =-.) Prof. King is currently a professor in the Mathematics Department at Colorado State University. Before, she was a professor at the University of Bremen; a Humboldt Fellow at the Technical University of Berlin, the University of Bonn, and the University of Osnabrück; a IRTA Postdoc Fellow at the National Institutes of Health; and a Ph.D. student at the University of Maryland. Please feel free to chat with her if you see her around virtual Günne. Photo credit: MFO / Petra Lein
Lecturer:Maithilee Kunda Fields: Artificial Intelligence/Cognitive Science
Content
While nearly 60 years of AI research on solving intelligence tests has yielded many techniques for many tests, we are still quite far from having an artificial agent that can “sit down and take” an intelligence test without specialized algorithms having been designed for that purpose. This course will discuss: 1) why intelligence tests are such a good challenge for AI; 2) a framework for artificial problem-solving agents with four components: a problem definition, input processing, domain knowledge, and a problem-solving strategy or procedure; 3) several types of agents from my own research that use visual-imagery-based strategies to solve problems from the well-known Raven\’s Progressive Matrices tests; and 4) ways in which an imagery-based agent could learn its domain knowledge, problem-solving strategies, and problem definition/input processing components from experience, instead of each being manually designed. We will also discuss implications of this work in understanding cultural differences in cognition, cultural and racial biases in the history of intelligence tests, and the worrying and still-prevalent phenomenon of stereotype threat.
Literature
Kunda, M. (2020). AI, visual imagery, and a case study on the challenges posed by human intelligence tests. Proceedings of the National Academy of Sciences, 117(47), 29390-29397. URL: https://doi.org/10.1073/pnas.1912335117
Kunda, M. (2019). Nonverbal task learning. In Proceedings of the Seventh Annual Conference on Advances in Cognitive Systems, 609-622. URL: https://cdn.vanderbilt.edu/vu-my/wp-content/uploads/sites/2127/2016/06/30080321/Kunda-2019-Nonverbal-task-learning.pdf
Hernández-Orallo, J., Martínez-Plumed, F., Schmid, U., Siebers, M., & Dowe, D. L. (2016). Computer models solving intelligence test problems: Progress and implications. Artificial Intelligence, 230, 74-107. URL: https://doi.org/10.1016/j.artint.2015.09.011
Evans, T. G. (1964, April). A heuristic program to solve geometric-analogy problems. In Proceedings of the April 21-23, 1964, spring joint computer conference (pp. 327-338). URL: https://doi.org/10.1145/1464122.1464156
Lecturer
Dr. Maithilee Kunda
Maithilee Kunda holds a B.S. in mathematics with computer science from MIT and a Ph.D. in computer science from Georgia Tech. She is currently an assistant professor of computer science and computer engineering at Vanderbilt University. Her work in artificial intelligence, in the area of cognitive systems, looks at how visual thinking contributes to learning and intelligent behavior, with a focus on applications for individuals on the autism spectrum. In 2016, she was recognized as a visionary on the MIT Technology Review’s annual list of 35 Innovators Under 35 for her work at the intersection of autism, AI, and visual thinking, and in 2020, her research on visuospatial cognitive assessment was featured on CBS 60 Minutes with Anderson Cooper, as part of a segment on, “Recruiting for talent on the autism spectrum.”
Lecturer:Barbara Hammer Fields: Machine Learning / AI safety
Content
Deep networks are prone to a number of attacks: data poisoning, backdoor attacks, adversarial examples, etc. In particular the latter have gained a lot of attention recently, since they display a shockingly different behavior of humans and machines, which can lead to severe threats in safety critical domains such as autonomous driving or automated border control. In the talk, I will give an overview about recent forms of attacks on deep neural networks, and have a glimpse on attempts of defenses such as adversarial training and further insights on what these findings tell us about the peculiarities of deep networks.
Literature
Chakraborty, Alam, Dey, Chattopadhyay, and Mukhopadhyay (2018). Adversarial Attacks and Defences: A Survey. arXiv:1810.00069
Machado, Silva, and Goldschmidt (2020). Adversarial Machine Learning in Image Classification: A Survey Towards the Defender’s Perspective. arXiv:2009.03728
Lecturer
Prof. Barbara Hammer
Barbara Hammer is a full Professor for Machine Learning at the CITEC Cluster at Bielefeld University, Germany. She received her Ph.D. in Computer Science in 1999 and her venia legendi (permission to teach) in 2003, both from the University of Osnabrueck, Germany, where she was head of an independent research group on the topic ‘Learning with Neural Methods on Structured Data’. In 2004, she accepted an offer for a professorship at Clausthal University of Technology, Germany, before moving to Bielefeld in 2010. Barbara\’s research interests cover theory and algorithms in machine learning and neural networks and their application for technical systems and the life sciences, including explainability, learning with drift, nonlinear dimensionality reduction, recursive models, and learning with non-standard data. Barbara has been chairing the IEEE CIS Technical Committee on Data Mining and Big Data Analytics, the IEEE CIS Technical Committee on Neural Networks, and the IEEE CIS Distinguished Lecturer Committee. She has been elected as member of the IEEE CIS Administrative Committee and the INNS Board. She is an associate editor of the IEEE Computational Intelligence Magazine, the IEEE TNNLS, and IEEE TPAMI. Currently, large parties of her work focusses on explainable machine learning for spatial-temporal data in her role as a PI of the ERC Synergy Grant Water-Futures.
Digital methods and machine learning are increasingly used in psychiatry. AI applications can facilitate the early detection and diagnosis of mental health problems, therapeutic practice is complemented by videoconferencing and texting, and methods from computational neuroscience are being applied to research in clinical psychiatry. The resulting digital and computational psychiatry has great potential to benefit patients. Apart from that, it may transform the way we think about diagnosis, treatment, and the very concept of a mental disorder. This raises philosophical and ethical questions. In this talk I will first present some practical and theoretical problems in psychiatry. I will then review how digital and computational methods, including AI, promise to contribute to solutions to these problems, and will discuss associated philosophical and ethical problems.
Literature
Fiske, A., Henningsen, P., & Buyx, A. (2019). Your Robot Therapist Will See You Now: Ethical Implications of Embodied Artificial Intelligence in Psychiatry, Psychology, and Psychotherapy. Journal of Medical Internet Research, 21(5), e13216. https://doi.org/10.2196/13216
Starke, G., De Clercq, E., Borgwardt, S., & Elger, B. S. (2020). Computing schizophrenia: Ethical challenges for machine learning in psychiatry. Psychological Medicine, 1–7. https://doi.org/10.1017/S0033291720001683
Uusitalo, S., Ma, J. T., & Arstila, V. (2020). Mapping out the philosophical questions of AI and clinical practice in diagnosing and treating mental disorders. Journal of Evaluation in Clinical Practice, 1–7. https://doi.org/10.1111/jep.13485
Wiese, W. (accepted). Von der KI-Ethik zur Bewusstseinsethik: Ethische Aspekte der Computational Psychiatry. [From the Ethics of AI to the Ethics of Consciousness: Ethical Aspects of Computational Psychiatry.] Psychiatrische Praxis.
Lecturer
Wanja Wiese received his PhD at Johannes Gutenberg University Mainz. His research focuses on consciousness and philosophical problems in cognitive science. He is editor-in-chief of the open-access journal Philosophy and the Mind Sciences (https://www.philosophymindscience.org).
Lecturer:Barbara Webb Fields: Computational Neuroscience, Robotics, AI
Content
Insect navigation has been a focus of behavioural study for many years, and provides a striking example of cognitive complexity in a miniature brain. We have used computational modelling to bridge the gap from behaviour to neural mechanisms by relating the computational requirements of navigational tasks to the type of computation offered by invertebrate brain circuits. We have shown that visual memory of multiple views could be acquired by associative learning in the mushroom body neuropil, and allow insects to recapitulate long routes. We have also proposed a circuit in the central complex neuropil that integrates sky compass and optic flow information on an outbound path and can thus steer the animal directly home. The models are strongly constrained by neuroanatomy, and are tested in realistic agent and robot simulations.
Literature
Webb, B. (2020). Robots with insect brains. Science, 368(6488), 244-245. Webb, B. (2019). The internal maps of insects. Journal of Experimental Biology, 222(Suppl 1).
Stone, T., Webb, B., Adden, A., Weddig, N. B., Honkanen, A., Templin, R., Wcislo, W., Scimeca, L., Warrant, E. & Heinze, S. (2017). An anatomically constrained model for path integration in the bee brain. Current Biology, 27(20), 3069-3085.
Ardin, P., Peng, F., Mangan, M., Lagogiannis, K., & Webb, B. (2016). Using an insect mushroom body circuit to encode route memory in complex natural environments. PLoS computational biology, 12(2), e1004683.
Lecturer
Barbara Webb joined the School of Informatics at the University of Edinburgh in May 2003. Previously she lectured at the University of Stirling (1999-2003), the University of Nottingham (1995-1998) and the University of Edinburgh (1993-1995). She received her Ph.D. (in Artificial Intelligence) from the University of Edinburgh in 1993, and her B.Sc. (in Psychology) from the University of Sydney in 1988. Her main research interest is in perceptual systems for the control of behaviour, through building computational and physical (robot) models of the hypothesised mechanisms. In particular she focuses on insect behaviours, as their smaller nervous systems may be easier to understand. Recent work includes study of some of the more complex capabilities of insects, including multimodal integration (in crickets and flies), navigation (in ants) and learning (in flies and maggots). She also has an interest in theoretical issues of methodology; in particular the problems of measurement, modeling and simulation.
Mechanoreception of pressure applied to the skin is the basis of the most direct perception of space. When the receptive field of a mechanoreceptor is stimulated, it will fire an action potential. The classical idea of space encoding was a labeled line code, with each mechanoreceptor representing a tiny area of skin. Increasing pressure to that point in space is encoded by an increasing action potential frequency. However: is this the most efficient way of encoding tactile stimuli? For the tiny nervous system of the leech, which is evolutionary optimized for a minimum number of cells and energy consumption, we found a different strategy: Each point in space on the skin is innervated by two mechanoreceptors with overlapping, spatially extended receptive fields. For each of these cells, the position of the stimulus in the receptive field AND the stimulus intensity influence two response features: the frequency AND the timing of the action potentials. Hence, for each individual cell, the representation of stimulus intensity in space is ambiguous. However, when the responses of only two cells are combined, the combination of stimulus location and intensity is encoded unambiguously in a much larger region than it would be possible with a labeled line code. Hence, the leech uses the smallest possible system for multiplexing: two cells representing two stimulus properties with two response features. This is good news not only for computational neuroscientists, who rejoice in a biological system implementing a minimal encoding strategy. Since the mechanoreceptors of leeches and humans share surprising similarities, this finding might also be relevant e.g. for the development of hand prosthesis providing sensory input.
Literature
Kretzberg J, Pirschel F, Fathiazar E and Hilgen G (2016) Encoding of Tactile Stimuli by Mechanoreceptors and Interneurons of the Medicinal Leech. Front. Physiol. 7:506. doi: 10.3389/fphys.2016.00506
Pirschel, F., Hilgen, G., Kretzberg, J. (2018) Effects of Touch Location and Intensity on Interneurons of the Leech Local Bend Network. Scientific Reports 8:3046. DOI:10.1038/s41598-018-21272-6
Pirschel, F., Kretzberg, J. (2016) Multiplexed Population Coding of Stimulus Properties by Leech Mechanosensory Cells. Journal of Neuroscience 36(13):3636 –3647. DOI:10.1523/JNEUROSCI.1753-15.2016
Saal HP, Bensmaia SJ (2014) Touch is a team effort: interplay of submodalities in cutaneous sensibility. Trends Neurosci 37:689–697. http://dx.doi.org/10.1016/j.tins.2014.08.012
Lecturer
Jutta Kretzberg studied computer science and biology at University of Bielefeld, Germany. In her PhD in biology she modelled neuronal responses in the fly visual system. As a postdoc in San Diego, California, she started to work also experimentally on the leech tactile system. In 2004, Jutta Kretzberg became a Junior Professor at University of Oldenburg, Germany, where she is now professor for computational neuroscience and head of the master’s program neuroscience. As a member of the cluster of excellence Hearing4all and having worked also on the vertebrate retina, her main research interest is neural coding in different sensory systems of vertebrates (including humans) and invertebrates. While juggling her family, teaching, research and administration duties, her favorite task is mentoring.
Lecturer:Viola Priesemann Fields: Networks, neural information processing, COVID
Content
We introduce the spreading dynamics of activity in neural networks, and then show how it fosters information processing.
Literature
Cramer et al., Nature Communications, 2020. https://www.nature.com/articles/s41467-020-16548-3
Contreras et al., 2020. https://arxiv.org/pdf/2011.11413
Contreras et al., Nature Communications, 2021. https://www.nature.com/articles/s41467-020-20699-8
Dehning et al., Science 2020. https://science.sciencemag.org/content/early/2020/05/14/science.abb9789
Wilting et al., Nature Communications, 2018. https://www.nature.com/articles/s41467-018-04725-4
Lecturer
Dr. Viola Priesemann
Dr. Viola Priesemann is a researcher at the Max Planck Institute for Dynamics and Self-Organization and teaches at the Georg-August University Göttingen. She studies spreading processes, self-organization and information processing in living and artificial networks. Since the COVID-19 outbreak, she has been studying the spread of SARS-CoV-2, quantified the effectiveness of interventions, and developed containment strategies. Viola Priesemann is co-author of several position papers (e.g. of the National Academy Leopoldina), Fellow of the Schiemann-Kolleg and member of the Cluster of Excellence “Multiscale Bioimaging” at the Campus Göttingen.
Affiliation: Max Planck Institute for Dynamics and Self-Organization Homepage:www.viola-priesemann.de
Modeling is pervasive in computing, even so there is no general definition of ‘computer modeling’. As some authors emphasise, each program embodies a model, though it may be implicit. For explicit modeling, the notion of modeling is normally taken over from the natural, social, cognitive or technical sciences relevant to the specific area of interest. Common to all models in computing is that they are ‘operational’ – when implemented and executed in computer programs, they become effective: they serve to enable or constrain human action and communication, or even to have direct impact on the real world.
Modeling inherently relies on ‘abstraction’ which involves reduction and decontextualisation. Human modellers – sometimes on their own, but more often working in teams and subject to collective interests – make choices in designing the model. These choices affect
the model base, i. e. which theories and concepts are suitable for modelling?
the model features, i.e. which objects, attributes, actions and relationships in the area of interested are relevant?
the model implementation, i. e. which technical platforms are appropriate and how are they used?
the methods for modeling, i. e. which techniques, tools and forms of organisation are used in design and implementation?
Design choices determine the basic properties of the operational system, when implemented and embedded in socio-technical environments: 1) its model-controlled behaviour, 2) its model- induced perception, and 3) the model-enabled human-computer interaction.
Thus, modellers have a scope for choice, which is characteristic for all design. Choice means freedom (within external constraints), and freedom inherently comes with responsibility. As designers, we cannot escape this responsibility, but we can accept, acknowledge and reflect it, making our choices transparent. We can develop value-based criteria for choices, communicate them to all stakeholders, discuss and agree on priorities and reach joint decisions.
In responsibility for design, technical quality of the operational system is important but not sufficient. Primary attention needs to be given to re-contextualization, i. e. to the embedding of the operational system in its socio-technical context. Therefore, responsible modellers cannot stay in their technical area of expertise only, but need to find ways for reaching out into the context and base their design decisions on careful anticipation of the operational system in use.
Lecturer
Prof. Dr. Christiane Floyd
Christiane Floyd is professor emerita for software engineering and honorary professor at the Technical University of Vienna. Floyd obtained her doctorate in Mathematics at the University of Vienna. She gained experience in computing as compiler developer at Siemens, Munich, as research associate and instructor at Stanford University, and as senior consultant for software development methods at Softlab, Munich. As head of the software engineering group at the Technical University of Berlin (1978-1991) and the University of Hamburg (1991-2008), she was the principal author of STEPS, a participatory and evolutionary approach to software development. Throughout her career, she pursued her interest in philosophical foundations of computing and had a strong concern for the responsible use of computing technology. Since 2006 she is committed to promoting the use of information and communication technologies for development in Ethiopia.
