Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Making spiking neurons more succinct with multi-compartment models

: Leugering, J.


Okandan, M. ; Association for Computing Machinery -ACM-:
Annual Neuro-Inspired Computational Elements Workshop, NICE 2020. Proceedings : Scheduled for Heidelberg, Germany, Postponed due to Covid-19 Pandemic, March, 2020
New York: ACM, 2020
ISBN: 978-1-4503-6123-1
ISBN: 978-1-4503-7718-8
Art. 7, 6 S.
Annual Neuro-Inspired Computational Elements Workshop (NICE) <2020, Heidelberg/cancelled>
Fraunhofer IIS ()

Spiking neurons consume energy for each spike they emit. Reducing the firing rate of each neuron - - without sacrificing relevant information content - - is therefore a critical constraint for energy efficient networks of spiking neurons in biology and neuromorphic hardware alike. The inherent complexity of biological neurons provides a possible mechanism to realize a good trade-off between these two conflicting objectives: multi-compartment neuron models can become selective to highly specific input patterns, and thus learn to produce informative yet sparse spiking codes. In this paper, I motivate the operation of a simplistic hierarchical neuron model by analogy to decision trees, show how they can be optimized using a modified version of the greedy decision tree learning rule, and analyze the results for a simple illustrative binary classification problem.