eprintid: 28385
rev_number: 13
eprint_status: archive
userid: 3631
dir: disk0/00/02/83/85
datestamp: 2020-06-10 15:38:00
lastmod: 2020-07-02 11:24:39
status_changed: 2020-06-10 15:38:00
type: doctoralThesis
metadata_visibility: show
creators_name: Kungl, Ákos Ferenc
title: Robust learning algorithms for spiking and rate-based neural networks
subjects: ddc-004
subjects: ddc-530
divisions: i-130700
adv_faculty: af-13
abstract: Inspired by the remarkable properties of the human brain, the fields of machine learning, computational neuroscience and neuromorphic engineering have achieved significant synergistic progress in the last decade. Powerful neural network models rooted in machine learning have been proposed as models for neuroscience and for applications in neuromorphic engineering. However, the aspect of robustness is often neglected in these models. Both biological and engineered substrates show diverse imperfections that deteriorate the performance of computation models or even prohibit their implementation. This thesis describes three projects aiming at implementing robust learning with local plasticity rules in neural networks. First, we demonstrate the advantages of neuromorphic computations in a pilot study on a prototype chip. Thereby, we quantify the speed and energy consumption of the system compared to a software simulation and show how on-chip learning contributes to the robustness of learning. Second, we present an implementation of spike-based Bayesian inference on accelerated neuromorphic hardware. The model copes, via learning, with the disruptive effects of the imperfect substrate and benefits from the acceleration. Finally, we present a robust model of deep reinforcement learning using local learning rules. It shows how backpropagation combined with neuromodulation could be implemented in a biologically plausible framework. The results contribute to the pursuit of robust and powerful learning networks for biological and neuromorphic substrates.
date: 2020
id_scheme: DOI
id_number: 10.11588/heidok.00028385
ppn_swb: 1703312724
own_urn: urn:nbn:de:bsz:16-heidok-283854
date_accepted: 2020-05-29
advisor: HASH(0x55a9a6343b28)
language: eng
bibsort: KUNGLAKOSFROBUSTLEAR2020
full_text_status: public
place_of_pub: Heidelberg
citation:   Kungl, Ákos Ferenc  (2020) Robust learning algorithms for spiking and rate-based neural networks.  [Dissertation]     
document_url: https://archiv.ub.uni-heidelberg.de/volltextserver/28385/1/Kungl_Akos_Ferenc_Doktorarbeit.pdf