arXiv subjects

Luke Taylor

Publications and source records attributed to Luke Taylor.

5 recordsLinked to original sources

arXiv2023

Addressing the speed-accuracy simulation trade-off for adaptive spiking neurons

The adaptive leaky integrate-and-fire (ALIF) model is fundamental within computational neuroscience and has been instrumental in studying our brains $\textit{in silico}$. Due to the sequential nature of simulating these neural models, a commonly faced issue is the speed-accuracy trade-off: either accurately simulate a neuron using a small discretisation time-step (DT), which is slow, or more quickly simulate a neuron using a larger DT and incur a loss in simulation accuracy. Here we provide a solution to this dilemma, by algorithmically reinterpreting the ALIF model, reducing the sequential simulation complexity and permitting a more efficient parallelisation on GPUs. We computationally validate our implementation to obtain over a $50\times$ training speedup using small DTs on synthetic benchmarks. We also obtained a comparable performance to the standard ALIF implementation on different supervised classification tasks - yet in a fraction of the training time. Lastly, we showcase how our model makes it possible to quickly and accurately fit real electrophysiological recordings of cortical neurons, where very fine sub-millisecond DTs are crucial for capturing exact spike timing.

cs.NE↗

arXiv2022

Robust and accelerated single-spike spiking neural network training with applicability to challenging temporal tasks

Spiking neural networks (SNNs), particularly the single-spike variant in which neurons spike at most once, are considerably more energy efficient than standard artificial neural networks (ANNs). However, single-spike SSNs are difficult to train due to their dynamic and non-differentiable nature, where current solutions are either slow or suffer from training instabilities. These networks have also been critiqued for their limited computational applicability such as being unsuitable for time-series datasets. We propose a new model for training single-spike SNNs which mitigates the aforementioned training issues and obtains competitive results across various image and neuromorphic datasets, with up to a $13.98\times$ training speedup and up to an $81\%$ reduction in spikes compared to the multi-spike SNN. Notably, our model performs on par with multi-spike SNNs in challenging tasks involving neuromorphic time-series datasets, demonstrating a broader computational role for single-spike SNNs than previously believed.

cs.NE↗

arXiv2017

A comparison between the Shooting and Finite-Difference Method in solving a Nonlinear Boundary Value Problem found in the context of light propagation

The shooting and finite-difference method are both numeric methods that approximate the solution of a BVP to a given accuracy. In this report both methods were implemented in Matlab and compared to each other on a BVP found in the context of light propagation in nonlinear dielectrics. It was observed that the finite-difference method is numerically more stable and converges faster than the shooting method.

math.NA↗

arXiv2017

A comparison between the Split Step Fourier and Finite-Difference method in analysing the soliton collision of a type of Nonlinear Schr\"odinger equation found in the context of optical pulses

In this report a type of Schr\"odinger Equation which is found in the context of optical pulses is analysed using the $\textit{Split Step}$ and $\textit{Finite Difference}$ method. The investigation shows interesting dynamics regarding certain values for parameter $S$ as well as a comparison between the two numeric schemes demonstrating the $\textit{Split Step}$ to be superior for this problem.

math.NA↗

arXiv2017

Improving Deep Learning using Generic Data Augmentation

Deep artificial neural networks require a large corpus of training data in order to effectively learn, where collection of such training data is often expensive and laborious. Data augmentation overcomes this issue by artificially inflating the training set with label preserving transformations. Recently there has been extensive use of generic data augmentation to improve Convolutional Neural Network (CNN) task performance. This study benchmarks various popular data augmentation schemes to allow researchers to make informed decisions as to which training methods are most appropriate for their data sets. Various geometric and photometric schemes are evaluated on a coarse-grained data set using a relatively simple CNN. Experimental results, run using 4-fold cross-validation and reported in terms of Top-1 and Top-5 accuracy, indicate that cropping in geometric augmentation significantly increases CNN task performance.

cs.LG↗