Papers by Nicolangelo Iannella
We derive an approximate analytical solution of a nonlinear cable equation describing the backpro... more We derive an approximate analytical solution of a nonlinear cable equation describing the backpropagation of action potentials in sparsely excitable dendrites with clusters of transiently activating, TTX-sensitive Na+ channels of low density, discretely distributed as point sources of transmembrane current along a continuous (non-segmented) passive cable structure. Each cluster or hot-spot, corresponding to a mesoscopic level description of Na+ ion channels, included known cumulative inactivation kinetics observed at the microscopic level. In such a reduced third-order system, the ‘recovery’ variable is an electrogenic sodium-pump and/or a Na+-Ca2+ exchanger imbedded in the passive membrane, and a high leakage conductance stabilizes the system. A nonlinear cable equation was used to investigate back-propagation and repetitive activity of action potentials, exhibiting characteristics of the modified Hodgkin-Huxley kinetics (in the presence of suprathreshold input). In particular, a time-dependent analytical solution was obtained through a perturbation expansion of the non-dimensional membrane potential (Φ) for all voltage dependent terms including the voltage dependent Na+ activation μ) and state-dependent inactivation (η) gating variables and then solving the resulting system of integral equations. It was shown that back-propagating action potentials attenuate in amplitude with the frequency following experimental findings and that the discrete and low-density distributions of transient Na+ channels along the cable structure contribute significantly to their discharge patterns. A major significance of integrative modelling is the provision of a continuous description of the non-dimensional membrane potential (Φ) as a function of position.
arXiv (Cornell University), Apr 8, 2012
Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and th... more Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and the formation of computational function in the brain. The classical model of STDP which considers the timing between pairs of pre-synaptic and post-synaptic spikes (p-STDP) is incapable of reproducing synaptic weight changes similar to those seen in biological experiments which investigate the effect of either higher order spike trains (e.g. triplet and quadruplet of spikes) [1]-[3], or, simultaneous effect of the rate and timing of spike pairs [4] on synaptic plasticity. In this paper, we firstly investigate synaptic weight changes using a p-STDP circuit [5] and show how it fails to reproduce the mentioned complex biological experiments. We then present a new STDP VLSI circuit which acts based on the timing among triplets of spikes (t-STDP) that is able to reproduce all the mentioned experimental results. We believe that our new STDP VLSI circuit improves upon previous circuits, whose learning capacity exceeds current designs due to its capability of mimicking the outcomes of biological experiments more closely; thus plays a significant role in future VLSI implementation of neuromorphic systems.
Multilayer perceptrons have received much attention in recent years due to their universal approx... more Multilayer perceptrons have received much attention in recent years due to their universal approximation capabilities. Normally, such models use real valued continuous signals, although they are loosely based on biological neuronal networks that encode signals using spike trains. Spiking neural networks are of interest both from a biological point of view and in terms of a method of robust signaling in particularly noisy or dif®cult environments. It is important to consider networks based on spike trains. A basic question that needs to be considered however, is what type of architecture can be used to provide universal function approximation capabilities in spiking networks? In this paper, we propose a spiking neural network architecture using both integrate-and-®re units as well as delays, that is capable of approximating a real valued function mapping to within a speci®ed degree of accuracy.
arXiv (Cornell University), Aug 21, 2011
We present new computational building blocks based on memristive devices. These blocks, can be us... more We present new computational building blocks based on memristive devices. These blocks, can be used to implement either supervised or unsupervised learning modules. This is achieved using a crosspoint architecture which is an efficient array implementation for nanoscale two-terminal memristive devices. Based on these blocks and an experimentally verified SPICE macromodel for the memristor, we demonstrate that firstly, the Spike-Timing-Dependent Plasticity (STDP) can be implemented by a single memristor device and secondly, a memristor-based competitive Hebbian learning through STDP using a 1 × 1000 synaptic network. This is achieved by adjusting the memristor's conductance values (weights) as a function of the timing difference between presynaptic and postsynaptic spikes. These implementations have a number of shortcomings due to the memristor's characteristics such as memory decay, highly nonlinear switching behaviour as a function of applied voltage/current, and functional uniformity. These shortcomings can be addressed by utilising a mixed gates that can be used in conjunction with the analogue behaviour for biomimetic computation. The digital implementations in this paper use in-situ computational capability of the memristor.
arXiv (Cornell University), Apr 8, 2012
Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and th... more Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and the formation of computational function in the brain. The classical model of STDP which considers the timing between pairs of pre-synaptic and post-synaptic spikes (p-STDP) is incapable of reproducing synaptic weight changes similar to those seen in biological experiments which investigate the effect of either higher order spike trains (e.g. triplet and quadruplet of spikes) [1]-[3], or, simultaneous effect of the rate and timing of spike pairs [4] on synaptic plasticity. In this paper, we firstly investigate synaptic weight changes using a p-STDP circuit [5] and show how it fails to reproduce the mentioned complex biological experiments. We then present a new STDP VLSI circuit which acts based on the timing among triplets of spikes (t-STDP) that is able to reproduce all the mentioned experimental results. We believe that our new STDP VLSI circuit improves upon previous circuits, whose learning capacity exceeds current designs due to its capability of mimicking the outcomes of biological experiments more closely; thus plays a significant role in future VLSI implementation of neuromorphic systems.
arXiv (Cornell University), Jan 15, 2015
The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the de... more The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the default framework for describing physical laws. Due to its fundamental importance, there have been various attempts to find the origin of this structure from more elementary principles. In this paper, we show how the Minkowski spacetime structure arises naturally from the geometrical properties of three dimensional space when modeled by Clifford geometric algebra of three dimensions Cℓ(ℜ 3). We find that a time-like dimension along with the three spatial dimensions, arise naturally, as well as four additional degrees of freedom that we identify with spin. Within this expanded eight-dimensional arena of spacetime, we find a generalisation of the invariant interval and the Lorentz transformations, with standard results returned as special cases. The value of this geometric approach is shown by the emergence of a fixed speed for light, the laws of special relativity and the form of Maxwell's equations, without recourse to any physical arguments.
ABSTRACT One of the main aspects of designing and operation prediction of any power system is to ... more ABSTRACT One of the main aspects of designing and operation prediction of any power system is to simulate its Electromagnetic Transient's (EMT) performance. Among the available power line models, Universal Line Model (ULM) is an efficient and accurate model for this purpose. There are random system parameters which affect the quantities of EMT output qualities. This paper presents a novel global sensitivity and uncertainty analysis techniques to determine the effect of these uncertain factors on the variation of the EMT outputs.
arXiv (Cornell University), Jan 15, 2015
The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the de... more The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the default framework for describing physical laws. Due to its fundamental importance, there have been various attempts to find the origin of this structure from more elementary principles. In this paper, we show how the Minkowski spacetime structure arises naturally from the geometrical properties of three dimensional space when modeled by Clifford geometric algebra of three dimensions Cℓ(ℜ 3). We find that a time-like dimension along with the three spatial dimensions, arise naturally, as well as four additional degrees of freedom that we identify with spin. Within this expanded eight-dimensional arena of spacetime, we find a generalisation of the invariant interval and the Lorentz transformations, with standard results returned as special cases. The value of this geometric approach is shown by the emergence of a fixed speed for light, the laws of special relativity and the form of Maxwell's equations, without recourse to any physical arguments.
arXiv (Cornell University), Nov 1, 2016
In this paper we develop a framework allowing a natural extension of the Lorentz transformations.... more In this paper we develop a framework allowing a natural extension of the Lorentz transformations. To begin, we show that by expanding conventional four-dimensional spacetime to eight-dimensions that a natural generalization is indeed obtained. We then find with these generalized coordinate transformations acting on Maxwell's equations that the electromagnetic field transformations are nevertheless unchanged. We find further, that if we assume the absence of magnetic monopoles, in accordance with Maxwell's theory, our generalized transformations are then restricted to be the conventional ones. While the conventional Lorentz transformations are indeed recovered from our framework, we nevertheless provide a new perspective into why the Lorentz transformations are constrained to be the conventional ones. Also, this generalized framework may assist in explaining several unresolved questions in electromagnetism as well as to be able to describe quasi magnetic monopoles found in spin-ice systems.
Journal of Physics Communications
The four dimensional spacetime continuum, as first conceived by Minkowski, has become the dominan... more The four dimensional spacetime continuum, as first conceived by Minkowski, has become the dominant framework within which to describe physical laws. In this paper, we show how this four-dimensional structure is a natural property of physical three-dimensional space, if modeled with Clifford geometric algebra C ℓ ( R 3 ) . We find that Minkowski spacetime can be embedded within a larger eight dimensional structure. This then allows a generalisation of the invariant interval and the Lorentz transformations. Also, with this geometric oriented approach the fixed speed of light, the laws of special relativity and a generalised form of Maxwell’s equations, arise naturally from the intrinsic properties of the algebra without recourse to physical arguments. We also find new insights into the nature of time, which can be described as two-dimensional. Some philosophical implications of this approach as it relates to the foundations of physical theories are also discussed.
In this paper, we present a new triplet based STDP VLSI implementation, based on a previous publi... more In this paper, we present a new triplet based STDP VLSI implementation, based on a previous published pair-based STDP circuit [1]. Simulation results illustrate that the proposed VLSI circuit can reproduce similar results to those observed in various physiological STDP experiments [2-4], while the traditional pair-based VLSI implementation fails to do so [2].
How sensory information from the external world is extracted, learnt, stored within, and retrieve... more How sensory information from the external world is extracted, learnt, stored within, and retrieved from the brain ultimately relies on both the strengths and patterning of connections converging onto neuronal dendrites and the nonlinear nature of the membrane. The underlying processes that permit this ultimately relies upon the non trivial interplay between the nonlinear nature of the dendrite, synapse location, and processes that mediate synaptic plasticity. Collectively, this interplay shapes both the response properties of neurons and their behavior within a neural circuit. Experimental and theoretical studies are starting to support the notion that the learning process encodes information through the formation of synaptic clusters or hotspots that share similar features. We have previously shown that spike timing-dependent plasticity (STDP) organizes the strengths of spatially distributed synaptic connections by forming a tessellation pattern of segregated synaptic efficacy clusters that effectively partitions the dendrite. We have called this patterning a dendritic mosaic. Furthermore, we have recently shown how the formation of such patterns is influenced by the degree balance admitted by STDP. Here, we report that altering the morphology of the dendritic tree can strongly influence the emergence of the dendritic mosaic.
<p>In <b>A</b>, we present the original data by Sjöström <a href="http:... more <p>In <b>A</b>, we present the original data by Sjöström <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0102601#pone.0102601-Sjstrm2" target="_blank">[59]</a> for comparison with our simulation results. In <b>B</b>, we display the outcome of Shouval's calcium based plasticity rule where changes are driven by normalized peak calcium. Note how the resulting plasticity profiles for and as a function of frequency lack a notable crossover point at 40 Hz. In <b>C</b> and <b>D</b>, again CaDP is used but this is driven by a modified calcium-dependent variable where the ratio between the peak calcium and the maximal integral over time of the calcium concentration profile is used to drive plastic change. In <b>C</b> and <b>D</b>, one can observe distance dependent changes in the plasticity outcomes for the pairing frequency protocol (as explained in text). In <b>C</b>, we see that the msec curve gives rise to only LTP, while for the msec there is an initial region that gives rise to LTD but then at a particular frequency there is a switch from LTD to LTP. Furthermore, note that the two curves crossover at 42 Hz. This is similar to Sjöström's experimental findings in <b>A</b>. In <b>D</b>, one can see a rightward shift of the msec curve where it inherits a small LTD component, however for the msec data there is a notable rightward shift of the LTD region leading to increase in frequency when LTD switches to LTP.</p
Following Minkowski's formulation of special relativity, it is generally accepted that we liv... more Following Minkowski's formulation of special relativity, it is generally accepted that we live in a four-dimensional world consisting of three space and one time dimension. Due to its fundamental importance, a variety of arguments have been proposed over the years attempting to derive this spacetime structure from underlying physical principles. In our approach, we show how Minkowski spacetime arises from the geometrical properties of three dimensional space. We demonstrate this through modeling physical space with Clifford's geometric algebra of three dimensions. We indeed find using this representation that a time-like dimension arises naturally within this space but also extends spacetime to eight dimensions through incorporating four spin degrees of freedom. This expanded arena of spacetime produces a generalized group of Lorentz transformations and provides a natural description of fundamental particles. Nearly all standard results are returned in this expanded structur...
The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the de... more The four dimensional spacetime continuum, as originally conceived by Minkowski, has become the default framework within which to describe physical laws. Due to its fundamental nature, there have been various attempts to derive this structure from more fundamental physical principles. In this paper, we show how the Minkowski spacetime structure arises directly from the geometrical properties of three dimensional space when modeled by Clifford geometric algebra of three dimensions Cℓ(ℜ 3). We find that a time-like dimension, as well as three spatial dimensions, arise naturally, as well as four additional degrees of freedom that we identify with spin. Within this expanded eightdimensional arena of spacetime, we find a generalisation of the invariant interval and the Lorentz transformations, with standard results returned as special cases. The power of this geometric approach is shown by the derivation of the fixed speed of light, the laws of special relativity and the form of Maxwell's equations, without any recourse to physical arguments. We also produce a unified treatment of energy-momentum and spin, as well as predicting a new class of physical effects and interactions.
<p>In (<b>A</b>) and (<b>C</b>), note how the illustrated plasticit... more <p>In (<b>A</b>) and (<b>C</b>), note how the illustrated plasticity outcomes for a single pre- and postsynaptic pairing using <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0102601#pone.0102601.e011" target="_blank">Eqn. (2)</a>, derived from the calcium control hypothesis using calcium based plasticity rule, changes as a function of distance. In particular, comparing (<b>A</b>) to (<b>C</b>) note the increase in the time constant of the LTD portion of the STDP window when the position from the source of spike initiation is increased. This behavior mimics the location-dependent nature of the STDP window as observed in (<b>C</b>) and (<b>D</b>) as presented in previous physiological experiment <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0102601#pone.0102601-Froemke1" target="_blank">[31]</a>. Parameters used were , , , , , , , , and . (<a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0102601#pone-0102601-g002" target="_blank">Figs 2B and 2D</a> were adapted from <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0102601#pone.0102601-Froemke1" target="_blank">[31]</a>).</p
2016 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS), 2016
This paper implements an established signal flow platform with its foundation derived from a syst... more This paper implements an established signal flow platform with its foundation derived from a system of nonlinear integral equations, characterizing the functional behavior of the signal that traverses from the photoreceptor to the ganglion cell in the vision processing architecture. While an increase in computational speed over the conventional method of solving a system of nonlinear ordinary differential equations (ODEs) has been confirmed for a single ganglion cell, the notion is extended to a retinal dual-pathway simulation which provides for a significantly improved adoption for organic mechanisms. There are various numerical methods in solving such a system which all have a bearing on both speed and error, and as such, systematic analyses using two common forms of integral solving methods are shown to improve the overall performance of simulating the extended pathway of the retinal model.
The 2012 International Joint Conference on Neural Networks (IJCNN), 2012
Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and th... more Spike-Timing Dependent Plasticity (STDP) is believed to play an important role in learning and the formation of computational function in the brain. The classical model of STDP which considers the timing between pairs of pre-synaptic and post-synaptic spikes (p-STDP) is incapable of reproducing synaptic weight changes similar to those seen in biological experiments which investigate the effect of either higher order spike trains (e.g. triplet and quadruplet of spikes) [1]-[3], or, simultaneous effect of the rate and timing of spike pairs [4] on synaptic plasticity. In this paper, we firstly investigate synaptic weight changes using a p-STDP circuit [5] and show how it fails to reproduce the mentioned complex biological experiments. We then present a new STDP VLSI circuit which acts based on the timing among triplets of spikes (t-STDP) that is able to reproduce all the mentioned experimental results. We believe that our new STDP VLSI circuit improves upon previous circuits, whose learning capacity exceeds current designs due to its capability of mimicking the outcomes of biological experiments more closely; thus plays a significant role in future VLSI implementation of neuromorphic systems.
PLoS ONE, 2014
Cortical circuits in the brain have long been recognised for their information processing capabil... more Cortical circuits in the brain have long been recognised for their information processing capabilities and have been studied both experimentally and theoretically via spiking neural networks. Neuromorphic engineers are primarily concerned with translating the computational capabilities of biological cortical circuits, using the Spiking Neural Network (SNN) paradigm, into in silico applications that can mimic the behaviour and capabilities of real biological circuits/systems. These capabilities include low power consumption, compactness, and relevant dynamics. In this paper, we propose a new accelerated-time circuit that has several advantages over its previous neuromorphic counterparts in terms of compactness, power consumption, and capability to mimic the outcomes of biological experiments. The presented circuit simulation results demonstrate that, in comparing the new circuit to previous published synaptic plasticity circuits, reduced silicon area and lower energy consumption for processing each spike is achieved. In addition, it can be tuned in order to closely mimic the outcomes of various spike timing-and rate-based synaptic plasticity experiments. The proposed circuit is also investigated and compared to other designs in terms of tolerance to mismatch and process variation. Monte Carlo simulation results show that the proposed design is much more stable than its previous counterparts in terms of vulnerability to transistor mismatch, which is a significant challenge in analog neuromorphic design. All these features make the proposed design an ideal circuit for use in large scale SNNs, which aim at implementing neuromorphic systems with an inherent capability that can adapt to a continuously changing environment, thus leading to systems with significant learning and computational abilities.
PloS one, 2014
Finding the rules underlying how axons of cortical neurons form neural circuits and modify their ... more Finding the rules underlying how axons of cortical neurons form neural circuits and modify their corresponding synaptic strength is the still subject of intense research. Experiments have shown that internal calcium concentration, and both the precise timing and temporal order of pre and postsynaptic action potentials, are important constituents governing whether the strength of a synapse located on the dendrite is increased or decreased. In particular, previous investigations focusing on spike timing-dependent plasticity (STDP) have typically observed an asymmetric temporal window governing changes in synaptic efficacy. Such a temporal window emphasizes that if a presynaptic spike, arriving at the synaptic terminal, precedes the generation of a postsynaptic action potential, then the synapse is potentiated; however if the temporal order is reversed, then depression occurs. Furthermore, recent experimental studies have now demonstrated that the temporal window also depends on the de...
Uploads
Papers by Nicolangelo Iannella