Abstract:This paper develops a new approach to estimate predicted class probabilities in deep Spiking Neural Networks (SNN) that encourages faster classification. The proposed approach utilizes the temporal separation between the first spikes generated by the output neurons to estimate the predicted class probabilities which are then used with cross entropy loss for training the network. This maximizes the separation between the first spikes generated by the neuron associated with the correct class and neurons associated with other classes. Higher classification performance is obtained by maximising the temporal separation, which also drives the correct class neuron to spike earlier in the simulation. As a consequence, the predicted class may be determined from the first spike in the output layer, leading to quicker classification. The sensitivity factor for each neuron in the network is estimated via error-backpropagation during training. Using Spike Timing Dependent Plasticity (STDP) regulated by the estimated sensitivity factors, the network weights are updated. It results that the learning method is termed as Temporal Separation Modulated Spike Timing Dependent Plasticity (TSM-STDP). On the benchmark MNIST dataset, the performance of TSM-STDP has been assessed, and the evaluation results are compared with those of other learning methods for SNNs. Additionally, a histogram of the output layer's first spikes demonstrated that the right class neurons spiked earlier in the simulation than other class neurons, enabling faster classification. On real-world Attention Deficit Hyperactivity Disorder (ADHD) detection dataset, the effectiveness of TSM-STDP has also been assessed and compared with other available approaches. The performance comparison results clearly show that TSM-STDP can achieve classification performance comparable to other existing learning algorithms on benchmark and real-world datasets while requiring less time for classification.

Selective classification considering time series characteristics for spiking neural networks

RSNN: Recurrent Spiking Neural Networks for Dynamic Spatial-Temporal Information Processing

Knowing When to Stop: Delay-Adaptive Spiking Neural Network Classifiers with Reliability Guarantees

Efficient Spiking Neural Networks with Sparse Selective Activation for Continual Learning

Fast and Accurate Classification with a Multi-Spike Learning Algorithm for Spiking Neurons.

A Supervised Multi-Spike Learning Algorithm for Spiking Neural Networks

Time series classification models based on nonlinear spiking neural P systems

Reservoir computing models based on spiking neural P systems for time series classification

Toward Efficient Processing and Learning with Spikes: New Approaches for Multispike Learning

Efficient and Effective Time-Series Forecasting with Spiking Neural Networks

Agreeing to Stop: Reliable Latency-Adaptive Decision Making via Ensembles of Spiking Neural Networks

Curriculum Design Helps Spiking Neural Networks to Classify Time Series

A Multi-Spiking Neural Network Learning Model for Data Classification.

Spike Attention Coding for Spiking Neural Networks.

SelectiveNet: A Deep Neural Network with an Integrated Reject Option

Learning to Classify Faster Using Spiking Neural Networks

Risk-Controlled Selective Prediction For Regression Deep Neural Network Models

A Novel Spiking Neural Network of Receptive Field Encoding with Groups of Neurons Decision

ReSpike: Residual Frames-based Hybrid Spiking Neural Networks for Efficient Action Recognition

Sparser spiking activity can be better: Feature Refine-and-Mask spiking neural network for event-based visual recognition

SEENN: Towards Temporal Spiking Early-Exit Neural Networks