Custom DNN using Reward Modulated Inverted STDP Learning for Temporal Pattern Recognition

TL;DR

This paper introduces a novel reward-modulated inverted STDP learning algorithm for temporal spike pattern recognition, enabling efficient detection in sparse event data with minimal training.

Contribution

It proposes a new algorithm combining reward-modulatory, Hebbian, and anti-Hebbian learning for temporal spike recognition in sparse datasets.

Findings

Effective recognition of spoken digit spike patterns.

Outperforms existing state-of-the-art methods.

Requires short training intervals.

Abstract

Temporal spike recognition plays a crucial role in various domains, including anomaly detection, keyword spotting and neuroscience. This paper presents a novel algorithm for efficient temporal spike pattern recognition on sparse event series data. The algorithm leverages a combination of reward-modulatory behavior, Hebbian and anti-Hebbian based learning methods to identify patterns in dynamic datasets with short intervals of training. The algorithm begins with a preprocessing step, where the input data is rationalized and translated to a feature-rich yet sparse spike time series data. Next, a linear feed forward spiking neural network processes this data to identify a trained pattern. Finally, the next layer performs a weighted check to ensure the correct pattern has been detected.To evaluate the performance of the proposed algorithm, it was trained on a complex dataset containing spoken digits with spike information and its output compared to state-of-the-art.