Professor Eleni Vasilaki
DPhil, CEng, MIET
Department of Computer Science
Chair of Bioinspired Machine Learning
Head of the Machine Learning research group
Member of Complex Systems Modelling research group
e.vasilaki@sheffield.ac.uk
+44 114 222 1822
+44 114 222 1822
Regent Court (DCS)
Full contact details
Professor Eleni Vasilaki
Department of Computer Science
Regent Court (DCS)
211 Portobello
Sheffield
S1 4DP
Department of Computer Science
Regent Court (DCS)
211 Portobello
Sheffield
S1 4DP
- Profile
-
Eleni graduated with a Bachelor’s degree in Informatics and Telecommunications and a Master’s degree in Microelectronics from the University of Athens, before taking her DPhil (PhD) at Sussex, in Computer Science and Artificial Intelligence.
From 2004 to 2006 she worked at the University of Bern and from 2007 to 2009 at the Swiss Federal Institute of Technology Lausanne (EPFL). In 2009 she joined the University of Sheffield as Lecturer, where she is Professor since 2016.
- Research interests
-
Bioinspired Machine Learning, Neuromorphic Computing, Computational Neuroscience
Eleni and her team take inspiration from biological principles to design novel, machine learning techniques, and in particular reinforcement learning and reservoir computing methods. They also work closely with material scientists and engineers to design hardware that computes in a brain-like manner.
- Publications
-
Journal articles
- Optimising network interactions through device agnostic models.. CoRR, abs/2401.07387.
- Editorial: Focus issue on energy-efficient neuromorphic devices, systems and algorithms.. Neuromorph. Comput. Eng., 3, 40201-40201.
- Editorial: Focus issue on energy-efficient neuromorphic devices, systems and algorithms. Neuromorphic Computing and Engineering, 3(4), 040201-040201.
- Machine learning using magnetic stochastic synapses.. Neuromorph. Comput. Eng., 3, 21001-21001.
- SpaRCe: Improved Learning of Reservoir Computing Systems Through Sparse Representations.. IEEE Trans. Neural Networks Learn. Syst., 34, 824-838.
- A perspective on physical reservoir computing with nanomagnetic devices. Applied Physics Letters, 122(4), 040501-040501.
- A robotic model of hippocampal reverse replay for reinforcement learning.. Bioinspir Biomim.
- Strategies discovery in the active allothetic place avoidance task. Scientific Reports, 12.
- Reservoir computing for temporal data classification using a dynamic solid electrolyte ZnO thin film transistor. Frontiers in Electronics, 3. View this article in WRRO
- EchoVPR: Echo State Networks for Visual Place Recognition. IEEE Robotics and Automation Letters, 1-1.
- Compensatory variability in network parameters enhances memory performance in the Drosophila mushroom body. Proceedings of the National Academy of Sciences, 118(49). View this article in WRRO
- Quantifying the Computational Capability of a Nanomagnetic Reservoir Computing Platform with Emergent Magnetization Dynamics. View this article in WRRO
- A robust model of stimulus-specific adaptation validated on neuromorphic hardware. Scientific Reports, 11(1).
- Neuromorphic computation with a single magnetic domain wall. Scientific Reports, 11(1). View this article in WRRO
- Voltage-controlled superparamagnetic ensembles for low-power reservoir computing. Applied Physics Letters, 118(20).
- Behavioral characteristics as potential biomarkers of the development and phenotype of epilepsy in a rat model of temporal lobe epilepsy. Scientific Reports, 11. View this article in WRRO
- Non-numerical strategies used by bees to solve numerical cognition tasks. Proceedings of the Royal Society B: Biological Sciences, 288(1945).
- A semi-supervised sparse K-Means algorithm. Pattern Recognition Letters, 142, 65-71. View this article in WRRO
- An empirical comparison between stochastic and deterministic centroid initialisation for K-means variations.. Mach. Learn., 110, 1975-2003.
- An alternative to backpropagation through time. Nature Machine Intelligence, 2(3), 155-156.
- Memristors - From In-Memory Computing, Deep Learning Acceleration, and Spiking Neural Networks to the Future of Neuromorphic and Bio-Inspired Computing.. Adv. Intell. Syst., 2, 2000085-2000085.
- Constitutive differences in glucocorticoid responsiveness are related to divergent spatial information processing abilities. Stress, 23(1), 37-49.
- Sodium nitroprusside prevents the detrimental effects of glucose on the neurovascular unit and behaviour in zebrafish. DMM Disease Models and Mechanisms, 12(9). View this article in WRRO
- A generalised framework for detailed classification of swimming paths inside the Morris Water Maze.. Scientific Reports, 8(1), 15089-15089. View this article in WRRO
- Abstract concept learning in a simple neural network inspired by the insect brain.. PLoS Computational Biology, 14(9). View this article in WRRO
- A computational model of the integration of landmarks and motion in the insect central complex.. PLoS ONE, 12(2). View this article in WRRO
- A Model for an Angular Velocity-Tuned Motion Detector Accounting for Deviations in the Corridor-Centering Response of the Bee. PLoS Computational Biology, 12(5). View this article in WRRO
- Emulating short-term synaptic dynamics with memristive devices. Scientific Reports, 6. View this article in WRRO
- Detailed classification of swimming paths in the Morris Water Maze: multiple strategies within one trial. Scientific Reports, 5. View this article in WRRO
- Adaptation of short-term plasticity parameters via error-driven learning may explain the correlation between activity-dependent synaptic properties, connectivity motifs and target specificity. Frontiers in Computational Neuroscience, 8. View this article in WRRO
- Measuring Symmetry, Asymmetry and Randomness in Neural Network Connectivity. PLoS ONE, 9(7). View this article in WRRO
- QSpike tools: a generic framework for parallel batch preprocessing of extracellular neuronal signals recorded by substrate microelectrode arrays. Frontiers in Neuroinformatics, 8, 26-26. View this article in WRRO
- Transient and steady-state selection in the striatal microcircuit. Frontiers in Computational Neuroscience, 7, 192-192. View this article in WRRO
- Emergence of Connectivity Motifs in Networks of Model Neurons with Short- and Long-Term Plastic Synapses. PLoS ONE, 9(1). View this article in WRRO
- Democratic population decisions result in robust policy-gradient learning: a parametric study with GPU simulations. PLoS One, 6(5), e18539. View this article in WRRO
- Connectivity reflects coding: a model of voltage-based STDP with homeostasis. Nat Neurosci, 13(3), 344-352.
- A biologically inspired dynamic model for vision..
- Spike-based reinforcement learning in continuous state and action space: when policy gradient methods fail. PLoS Comput Biol, 5(12), e1000586. View this article in WRRO
- Adaptive gain modulation in V1 explains contextual modifications during bisection learning. PLoS Comput Biol, 5(12), e1000617. View this article in WRRO
- Modeling plasticity across different time scales: the TagTriC model. BMC Neuroscience, 10(Suppl 1), P192-P192.
- Stimulus sampling as an exploration mechanism for fast reinforcement learning. Biol Cybern, 100(4), 319-330.
- Learning flexible sensori-motor mappings in a complex network. Biol Cybern, 100(2), 147-158.
- Tag-trigger-consolidation: a model of early and late long-term-potentiation and depression. PLoS Comput Biol, 4(12), e1000248. View this article in WRRO
- Perceptual learning via modification of cortical top-down signals.. PLoS Comput Biol, 3(8), e165. View this article in WRRO
- Some optimal stochastic control problems in neuroscience - A review. MOD PHYS LETT B, 18(21-22), 1067-1085.
- Towards a Spatio-Temporal Album. WSEAS Trans. on Systems, 2(4), 941-947.
- Temporal album. IEEE Trans Neural Netw, 14(2), 439-443.
- Reconfigurable reservoir computing in a magnetic metamaterial. Communications Physics, 6(1).
- Modelling novelty detection in the thalamocortical loop. PLOS Computational Biology, 19(5), e1009616-e1009616.
- Quantifying the Computational Capability of a Nanomagnetic Reservoir Computing Platform with Emergent Magnetisation Dynamics. Nanotechnology.
- Signal neutrality, scalar property, and collapsing boundaries as consequences of a learned multi-timescale strategy. PLOS Computational Biology, 18(8), e1009393-e1009393.
- Dynamically‐Driven Emergence in a Nanomagnetic System. Advanced Functional Materials, 2008389-2008389.
- A Robotic Model of Hippocampal Reverse Replay for Reinforcement Learning.
- Analysis of behaviour in the Active Allothetic Place Avoidance task based on cluster analysis of the rat movement motifs.
- Decision-making and action selection in insects: inspiration from vertebrate-based theories. Frontiers in Behavioral Neuroscience, 9. View this article in WRRO
- Correction: Spike-Based Reinforcement Learning in Continuous State and Action Space: When Policy Gradient Methods Fail. PLoS Computational Biology, 5(12).
- Connectivity reflects coding: A model of voltage-based spike-timing-dependent-plasticity with homeostasis. Nature Precedings.
- Memristors—From In‐Memory Computing, Deep Learning Acceleration, and Spiking Neural Networks to the Future of Neuromorphic and Bio‐Inspired Computing. Advanced Intelligent Systems, 2000085-2000085.
- An empirical comparison between stochastic and deterministic centroid
initialisation for K-Means variations. View this article in WRRO
- SpaRCe: Sparse reservoir computing. View this article in WRRO
- Exploiting Multiple Timescales in Hierarchical Echo State Networks. Machine Learning.
- EchoVPR: Echo State Networks for Visual Place Recognition.
Conference proceedings papers
- A delay system reservoir based on a nano-ionic Solid Electrolyte FET*. 2023 IEEE Nanotechnology Materials and Devices Conference (NMDC), 22 October 2023 - 25 October 2023.
- Reservoir Computing Based on a Solid Electrolyte ZnO TFT: An Attractive Platform for Flexible Edge Computing. 2023 IEEE International Flexible Electronics Technology Conference (IFETC), 13 August 2023 - 16 August 2023.
- From stochasticity to functionality: harnessing magnetic domain walls for probabilistic and neuromorphic computing. Spintronics XIV, 1 August 2021 - 5 August 2021.
- Fast reverse replays of recent spatiotemporal trajectories in a robotic hippocampal model. Biomimetic and Biohybrid Systems : 9th International Conference, Living Machines 2020, Freiburg, Germany, July 28–30, 2020, Proceedings (pp 390-401). Freiburg, Germany, 28 July 2020 - 30 July 2020.
- Robots that imagine – can hippocampal replay be utilized for robotic mnemonics?. Biomimetic and Biohybrid Systems (pp 277-286). Nara, Japan, 9 July 2019 - 12 July 2019. View this article in WRRO
- An Inexpensive Flying Robot Design for Embodied Robotics Research,. 2017 International Joint Conference on Neural Networks (IJCNN) (pp 4171-4178), 15 May 2017 - 19 May 2017. View this article in WRRO
- Bio-Inspired Visual Navigation for a Quadcopter using Optic Flow. AIAA Infotech @ Aerospace View this article in WRRO
- Highly scalable parallel processing of extracellular recordings of Multielectrode Arrays. 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 25 August 2015 - 29 August 2015. View this article in WRRO
- A neural model of the optomotor system accounts for ordered responses to decreasing stimulus spatial frequencies. BMC Neuroscience, Vol. 16(S1)
- QSpikeTools: An open source toolbox for parallel batch processing of extracellular neuronal signals recorded by substrate microelectrode arrays. 2014 International Conference on Electrical Engineering and Information & Communication Technology, 10 April 2014 - 12 April 2014.
- A Web-Based Framework for Semi-Online Parallel Processing of Extracellular Neuronal Signals Recorded by Microelectrode Arrays. MEA Meeting 2014, 9th International Meeting on Substrate-Integrated Microelectrode Arrays, 1 July 2014 - 4 July 2014.
- Memristors as synapse emulators in the context of event-based computation. IEEE International Symposium on Circuits and Systems (ISCAS)
- QSpikeTools: An Open Source Toolbox for Parallel Batch Processing of Extracellular Neuronal Signals Recorded. 2014 1ST INTERNATIONAL CONFERENCE ON ELECTRICAL ENGINEERING AND INFORMATION & COMMUNICATION TECHNOLOGY (ICEEICT 2014)
- The green brain project - Developing a neuromimetic robotic honeybee. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 8064 LNAI (pp 362-363)
- INCREASED COGNITIVE FUNCTIONING IN MANIFEST HD-EMPIRICAL EVIDENCE AND COMPUTATIONAL MODELLING. JOURNAL OF NEUROLOGY NEUROSURGERY AND PSYCHIATRY, Vol. 83 (pp A37-A38)
- How Degrading Networks Can Increase Cognitive Functions. ICANN (1), Vol. 7552 (pp 185-192)
- Emergence of Connectivity Patterns from Long-Term and Short-Term Plasticities. ICANN (1), Vol. 7552 (pp 193-200)
- How degrading networks can increase select cognitive functions. Computational Neuroscience and Neurotechnology Bernstein Conference 2011. Freiburg, Germany, 4 October 2011.
- Why is connectivity in barrel cortex different from that in visual cortex? - A plasticity model. COSYNE Computational and Systems Neuroscience. Salt Lake City, Utah, USA., 25 February 2010.
- Pattern sampling-based reinforcement learning via homeostatic plasticity and Hebbian tagging. Society for Neuroscience. San Diego, USA.
- Hebbian reinforcement learning with stochastic binary synapses follows the reward gradient. The Fifteenth Annual Computational Neuroscience Meeting, Edinbourgh, UK.
- Temporal Processing with Volatile Memristors. IEEE International Symposium on Circuits and Systems (ISCAS) 2013. Beijing, China, 19 May 2013 - 23 May 2013.
- Do synaptic dynamics and STDP govern connectivity motifs?. Computational Neuroscience and Neurotechnology Bernstein Conference 2011. Freiburg, Germany, 4 October 2011.
- Spike-based reinforcement in continuous state and action space. Multidisciplinary Symposium on Reinforcement Learning. Montreal, Quebec, Canada., 18 June 2009.
- Spike-based reinforcement learning of navigation. The Seventeenth Annual Computational Neuroscience Meeting. Portland, Oregon, USA, 19 July 2008.
- Spike-based reinforcement learning of navigation. The Seventeenth Annual Computational Neuroscience Meeting. Portland, Oregon, USA, 19 July 2008.
- A unified voltage-based model for STDP, LTP and LTD. COSYNE Computational and Systems Neuroscience. Salt Lake City, Utah, USA., 22 February 2007.
- Learning and forgetting visuo-motor associations with a multi-layer neural network. Brain Inspired Cognitive systems. Molyvos, Island of Lesvos, Greece, 12 October 2006.
- Perceptual learning by modifying top-down connections to V1. The Fifteenth Annual Computational Neuroscience Meeting. Edinbourgh, UK
- A Hebbian reinforcement learning algorithm reproducing monkey performances in visuo-motor learning task. COSYNE Computational and Systems Neuroscience. Salt Lake City, Utah.
- Processing Static Visual Information with IF-networks. Proceedings of the 12th Danish Conference on Pattern Recognition and Image Analysis, 1 August 2003.
- Comparison of Feedforward (TDRBF) and Generative (TDRGBN) Network for Gesture Based Control. Lecture Notes in Artificial Intelligence, Vol 2298
- Analysis Of Phosphorus Diffusion In the Polysilicon/Oxide/Silicon System Under Oxidizing Conditions For Profile Simulation. 194th Meeting, Electrochemical Society. Boston, MA.
- Is Epicurus the father of Reinforcement Learning?. Sheffield Machine Learning Retreat 2017 View this article in WRRO
Software / Code
- RodentDataAnalytics/mwm-ml-gen: Version 4.0.3-beta.
- Data and Code for: Non-numerical strategies used by bees to solve numerical cognition tasks.
Preprints
- Optimising network interactions through device agnostic models, arXiv.
- Machine learning using magnetic stochastic synapses.
- A perspective on physical reservoir computing with nanomagnetic devices, arXiv.
- Adaptive Programmable Networks for In Materia Neuromorphic Computing, Research Square Platform LLC.
- Adaptive Programmable Networks for In Materia Neuromorphic Computing.
- Reservoir Computing with Emergent Dynamics in a Magnetic Metamaterial, Research Square Platform LLC.
- Reservoir Computing with Emergent Dynamics in a Magnetic Metamaterial.
- Signal neutrality, scalar property, and collapsing boundaries as consequences of a learned multi-timescale strategy, Cold Spring Harbor Laboratory.
- A Robotic Model of Hippocampal Reverse Replay for Reinforcement Learning, arXiv.
- Exploiting Multiple Timescales in Hierarchical Echo State Networks, arXiv.
- Memristors -- from In-memory computing, Deep Learning Acceleration, Spiking Neural Networks, to the Future of Neuromorphic and Bio-inspired Computing, arXiv.
- A semi-supervised sparse K-Means algorithm, arXiv.
- SpaRCe: Improved Learning of Reservoir Computing Systems through Sparse Representations, arXiv.
- An empirical comparison between stochastic and deterministic centroid initialisation for K-Means variations, arXiv.
- Sodium nitroprusside prevents the detrimental effects of glucose on the neurovascular unit and behaviour in zebrafish, Cold Spring Harbor Laboratory.
- Constitutive differences in glucocorticoid responsiveness are related to divergent spatial information processing abilities, Cold Spring Harbor Laboratory.
- A generalised framework for detailed classification of swimming paths inside the Morris Water Maze, arXiv.
- Is Epicurus the father of Reinforcement Learning?, arXiv.
- Emulating short-term synaptic dynamics with memristive devices, arXiv.
- Emergence of Connectivity Motifs in Networks of Model Neurons with Short- and Long-term Plastic Synapses, arXiv.
- Modelling novelty detection in the thalamocortical loop.
- Neuromorphic Computation With a Single Magnetic Domain Wall.
- Compensatory variability in network parameters enhances memory performance in the Drosophila mushroom body. View this article in WRRO
- Learning sparsity in reservoir computing through a novel bio-inspired
algorithm. View this article in WRRO
- Detection of multiple and overlapping bidirectional communities within
large, directed and weighted networks of neurons. View this article in WRRO
- Emulating long-term synaptic dynamics with memristive devices. View this article in WRRO
- Optimising network interactions through device agnostic models.. CoRR, abs/2401.07387.
- Grants
-
Current grants
- CausalXRL: Causal eXplanations in Reinforcement Learning, EPSRC, 02/2021 - 01/2024, £309,915, as PI
- MARCH: Magnetic Architectures for Reservoir Computing Hardware, EPSRC, 02/2021 - 07/2024, £936,815, as Co-PI
- ActiveAI - active learning and selective attention for robust, transparent and efficient AI, EPSRC, 11/2019 - 10/2024, £953,584, as Co-PI
- Modeling probabilistic reinforcement learning and variable behaviour in the fruit fly Drosophila melanogaster, Google, 05/2018 - 07/2024, £50,769, as PI
Previous grants
- From Stochasticity to Functionality: Probabilistic Computation with Magnetic Nanowires, EPSRC, 04/2019 - 11/2023, £755,424, as Co-PI
- Alexa Fellowship, Amazon, 08/2018 - 08/2021, £73,000, as Co-PI
- Brains on Board: Neuromorphic Control of Flying Robots, EPSRC, 12/2016 - 12/2021, £2,128,934, as Co-P
- From synaptic plasticity to cortical neuronal networks emergent behaviour, Royal Society, 06/2010 - 05/2012, £8,450, as PI
- The cortical representation of low-probability stimuli and its neuromorphic implementation, The Wellcome Trust, 09/2016 - 04/2021, £246,222, as PI
- NAMASEN: Neuroelectronics and nanotechnology - towards a Multidisciplinary Approach for the Science and Engineering of Neuronal Networks, EC FP7, 10/2011 - 09/2015, £215,509, as PI
- Green Brain: Computational Modelling of the Honeybee Brain, EPSRC, 03/2013 - 08/2016, £660,561, as Co-PI
- Memristive Dynamics, EPSRC, 05/2012 - 02/2013, £17,271, as PI
- Professional activities and memberships
-
Inge Strauch Visiting Professor by the University of Zurich (1 September 2021-28 February 2022).
One of the Academic editors of Scientific Reports, Nature Publishing Group, PLOS ONE and PeerJ.
Advisory Board Member for the research Centre For Emergent Algorithmic Intelligence at the University of Mainz, Germany, since 2019.
More than 50 invited talks, including international leading Institutions, e.g. IIT Roorkee, India (2020), Nature Conference for Neuromorphic Computing, Beijing, China (2019), European Institute for Theoretical Neuroscience (EITN), Paris (2018), Beijing Institute of Technology, China (2017), University of Oxford, UK (2016), Imperial College, UK (2015), University La Sapienza of Rome, Italy (2014), ETH Zurich & University of Zurich, Switzerland (2013), Columbia University, USA (2013) and keynote talks (e.g. International Multi-Conference on Artificial Intelligence Technology (M-CAIT2021).