Reservoir Computing and Self-Organized Neural Hierarchies

  • There is a growing understanding that machine learning architectures have to be much bigger and more complex to approach any intelligent behavior. There is also a growing understanding that purely supervised learning is inadequate to train such systems. A recent paradigm of artificial recurrent neural network (RNN) training under the umbrella-name Reservoir Computing (RC) demonstrated that training big recurrent networks (the reservoirs) differently than supervised readouts from them is often better. It started with Echo State Networks (ESNs) and Liquid State Machines ten years ago where the reservoir was generated randomly and only linear readouts from it were trained. Rather surprisingly, such simply and fast trained ESNs outperformed classical fully-trained RNNs in many tasks. While full supervised training of RNNs is problematic, intuitively there should also be something better than a random network. In recent years RC became a vivid research field extending the initial paradigm from fixed random reservoir and trained output into using different methods for training the reservoir and the readout. In this thesis we overview existing and investigate new alternatives to the classical supervised training of RNNs and their hierarchies. First we present a taxonomy and a systematic overview of the RNN training approaches under the RC umbrella. Second, we propose and investigate the use of two different neural network models for the reservoirs together with several unsupervised adaptation techniques, as well as unsupervisedly layer-wise trained deep hierarchies of such models. We rigorously empirically test the proposed methods on two temporal pattern recognition datasets, comparing it to the classical reservoir computing state of art.

Download full text

Cite this publication

  • Export Bibtex
  • Export RIS

Citable URL (?):

Search for this publication

Search Google Scholar Search Catalog of German National Library Search OCLC WorldCat Search Catalog of GBV Common Library Network Search Catalog of Jacobs University Library Search Bielefeld Academic Search Engine
Meta data
Publishing Institution:IRC-Library, Information Resource Center der Jacobs University Bremen
Granting Institution:Jacobs Univ.
Author:Mantas Lukosevicius
Referee:Herbert Jaeger, Helge Ritter, Mathias Bode
Advisor:Herbert Jaeger
Persistent Identifier (URN):urn:nbn:de:101:1-201306047929
Document Type:PhD Thesis
Language:English
Date of Successful Oral Defense:2011/12/01
Date of First Publication:2012/11/04
PhD Degree:Computer Science
School:SES School of Engineering and Science
Library of Congress Classification:Q Science / QA Mathematics (incl. computer science) / QA71-90 Instruments and machines / QA75.5-76.95 Electronic computers. Computer science / QA76.87 Neural computers. Neural networks
Call No:Thesis 2011/63

$Rev: 13581 $