Simeon Bamford PhD

Simeon Bamford


Research Interests

I am interested in how neural networks rewire themselves during development, so that for example brain cells which are connected to each other when you're a baby might end up connected to different cells later on. I'm using as a prime example how the connections from light-sensitive cells on the back of the eye organise themselves in the same order as they project to different areas of the brain. I want to create silicon chips with circuitry which behaves a bit like brain cells, and I want to use these artificial neurons to demonstrate this rewiring behaviour. Silicon chips designed in this way might ultimately lead to better designs of computers, and I'm interested in how models systems like the ones I'm creating could be scaled up into useable systems.

Publications:
2012
  Silicon synapses self-correct for both mismatch and design inhomogeneities
Bamford, S, Murray, A & Willshaw, D 2012, 'Silicon synapses self-correct for both mismatch and design inhomogeneities' Electronics Letters, vol 48, pp. 360-361.
General Information
Organisations: Neuroinformatics DTC.
Authors: Bamford, Simeon, Murray, Alan & Willshaw, David.
Number of pages: 2
Pages: 360-361
Publication Date: 2012
Publication Information
Category: Article
Journal: Electronics Letters
Volume: 48
ISSN: 0013-5194
Original Language: English
  Spike-timing-dependent plasticity with weight dependence evoked from physical constraints
Bamford, S, Murray, AF & Willshaw, DJ 2012, 'Spike-timing-dependent plasticity with weight dependence evoked from physical constraints' IEEE Transactions on Biomedical Circuits and Systems, vol 6, no. 4, pp. 385-98. DOI: 10.1109/TBCAS.2012.2184285
Analogue and mixed-signal VLSI implementations of Spike-Timing-Dependent Plasticity (STDP) are reviewed. A circuit is presented with a compact implementation of STDP suitable for parallel integration in large synaptic arrays. In contrast to previously published circuits, it uses the limitations of the silicon substrate to achieve various forms and degrees of weight dependence of STDP. It also uses reverse-biased transistors to reduce leakage from a capacitance representing weight. Chip results are presented showing: various ways in which the learning rule may be shaped; how synaptic weights may retain some indication of their learned values over periods of minutes; and how distributions of weights for synapses convergent on single neurons may shift between more or less extreme bimodality according to the strength of correlational cues in their inputs.
General Information
Organisations: Neuroinformatics DTC.
Authors: Bamford, Simeon, Murray, Alan F & Willshaw, David J.
Number of pages: 14
Pages: 385-98
Publication Date: Aug 2012
Publication Information
Category: Article
Journal: IEEE Transactions on Biomedical Circuits and Systems
Volume: 6
Issue number: 4
Original Language: English
DOIs: https://doi.org/10.1109/TBCAS.2012.2184285
2010
  Large Developing Receptive Fields Using a Distributed and Locally Reprogrammable Address-Event Receiver
Bamford, S, Murray, A & Willshaw, DJ 2010, 'Large Developing Receptive Fields Using a Distributed and Locally Reprogrammable Address-Event Receiver' Ieee Transactions on Neural Networks, vol 21, no. 2, pp. 286-304. DOI: 10.1109/TNN.2009.2036912
A distributed and locally reprogrammable address-event receiver has been designed, in which incoming address-events are monitored simultaneously by all synapses, allowing for arbitrarily large axonal fan-out without reducing channel capacity. Synapses can change the address of their presynaptic neuron, allowing the distributed implementation of a biologically realistic learning rule, with both synapse formation and elimination ( synaptic rewiring). Probabilistic synapse formation leads to topographic map development, made possible by a cross-chip current-mode calculation of Euclidean distance. As well as synaptic plasticity in rewiring, synapses change weights using a competitive Hebbian learning rule (spike-timing-dependent plasticity). The weight plasticity allows receptive fields to be modified based on spatio-temporal correlations in the inputs, and the rewiring plasticity allows these modifications to become embedded in the network topology.
General Information
Organisations: Institute for Adaptive and Neural Computation .
Authors: Bamford, Simeon, Murray, Alan & Willshaw, D. J..
Keywords: (Address–event representation (AER) , Euclidean distance , neural network architecture , neural network hardware , neuromorphic very large scale integration (VLSI) , synapse elimination , synapse formation , synaptic rewiring, topographic map, , , . )
Number of pages: 19
Pages: 286-304
Publication Date: Feb 2010
Publication Information
Category: Article
Journal: Ieee Transactions on Neural Networks
Volume: 21
Issue number: 2
ISSN: 1045-9227
Original Language: English
DOIs: https://doi.org/10.1109/TNN.2009.2036912
  Synaptic rewiring for topographic mapping and receptive field development
Bamford, S, Murray, A & Willshaw, DJ 2010, 'Synaptic rewiring for topographic mapping and receptive field development' Neural Networks, vol 23, no. 4, pp. 517-527. DOI: 10.1016/j.neunet.2010.01.005
A model of topographic map refinement is presented which combines both weight plasticity and the formation and elimination of synapses, as well as both activity-dependent and activity-independent processes. The question of whether an activity-dependent process can refine a mapping created by an activity-independent process is addressed statistically. A new method of evaluating the quality of topographic projections is presented which allows independent consideration of the development of the centres and spatial variances of receptive fields for a projection. Synapse formation and elimination embed in the network topology changes in the weight distributions of synapses due to the activity-dependent learning rule used (spike-timing-dependent plasticity). In this model, the spatial variance of receptive fields can be reduced by an activity-dependent mechanism with or without spatially correlated inputs, but the accuracy of receptive field centres will not necessarily improve when synapses are formed based on distributions with on-average perfect topography.
General Information
Organisations: Institute for Adaptive and Neural Computation .
Authors: Bamford, Simeon, Murray, Alan & Willshaw, D. J..
Keywords: (Synapse formation, Synapse elimination, Synaptic rewiring, Synaptic plasticity, Spike-timing-dependent plasticity (STDP), Activity dependent, Activity independent, Integrate-and-fire, Receptive field, Topographic map, Mapping , Map development , Ocular dominance , Topographic refinement, , , . )
Number of pages: 11
Pages: 517-527
Publication Date: May 2010
Publication Information
Category: Article
Journal: Neural Networks
Volume: 23
Issue number: 4
ISSN: 0893-6080
Original Language: English
DOIs: https://doi.org/10.1016/j.neunet.2010.01.005
2008
  Synaptic Rewiring for Topographic Map Formation
Bamford, S, Murray, A & Willshaw, DJ 2008, Synaptic Rewiring for Topographic Map Formation. in International Conference on Artificial Neural Networks (ICANN). pp. 218-227. DOI: 10.1007/978-3-540-87559-8_23
A model of topographic map development is presented which combines both weight plasticity and the formation and elimination of synapses as well as both activity-dependent and -independent processes. We statistically address the question of whether an activity-dependent process can refine a mapping created by an activity-independent process. A new method of evaluating the quality of topographic projections is presented which allows independent consideration of the development of a projection’s preferred locations and variance. Synapse formation and elimination embed in the network topology changes in the weight distributions of synapses due to the activity-dependent learning rule used (spike-timing-dependent plasticity). In this model, variance of a projection can be reduced by an activity dependent mechanism with or without spatially correlated inputs, but the accuracy of preferred locations will not necessarily improve when synapses are formed based on distributions with on-average perfect topography.
Artificial Neural Networks - ICANN 2008 Artificial Neural Networks - ICANN
General Information
Organisations: Neuroinformatics DTC.
Authors: Bamford, Simeon, Murray, Alan & Willshaw, D. J..
Number of pages: 10
Pages: 218-227
Publication Date: 3 Sep 2008
Publication Information
Category: Conference contribution
Original Language: English
DOIs: https://doi.org/10.1007/978-3-540-87559-8_23
  Large Developing Axonal Arbors Using a Distributed and Locally-Reprogrammable Address-Event Receiver
Bamford, S, Murray, A & Willshaw, DJ 2008, Large Developing Axonal Arbors Using a Distributed and Locally-Reprogrammable Address-Event Receiver. in IEEE International Joint Conference on Neural Networks (IJCNN). pp. 1464-1471. DOI: 10.1109/IJCNN.2008.4633990
We have designed a distributed and locally reprogrammable address event receiver. Incoming address-events are monitored simultaneously by all synapses, allowing for arbitrarily large axonal fan-out without reducing channel capacity. Synapses can change input address, allowing neurons to implement a biologically realistic learning rule locally, with both synapse formation and elimination.
General Information
Organisations: Neuroinformatics DTC.
Authors: Bamford, Simeon, Murray, Alan & Willshaw, D. J..
Number of pages: 8
Pages: 1464-1471
Publication Date: 1 Jun 2008
Publication Information
Category: Conference contribution
Original Language: English
DOIs: https://doi.org/10.1109/IJCNN.2008.4633990

Projects:
Synaptic rewiring in neuromorphic VLSI for topographic map formation (PhD)