PP-2004-16: Nonmonotonic Inferences and Neural Networks

PP-2004-16: Blutner, Reinhard (2004) Nonmonotonic Inferences and Neural Networks. [Report]

Preview	Text (Full Text) PP-2004-16.text.pdf Download (402kB) | Preview
	Text (Abstract) PP-2004-16.abstract.txt Download (1kB)

Abstract

There is a gap between two different modes of computation: the
symbolic mode and the subsymbolic (neuron-like) mode. The aim of this
paper is to overcome this gap by viewing symbolism as a high-level
description of the properties of (a class of) neural networks.
Combining methods of algebraic semantics and nonmonotonic logic, the
possibility of integrating both modes of viewing cognition is
demonstrated. The main results are (a) that certain activities of
connectionist networks can be interpreted as nonmonotonic inferences,
and (b) that there is a strict correspondence between the coding of
knowledge in Hopfield networks and the knowledge representation in
weight-annotated Poole systems. These results show the usefulness of
nonmonotonic logic as a descriptive and analytic tool for analyzing
emerging properties of connectionist networks. Assuming an exponential
development of the weight function, the present account relates to
optimality theory - a general framework that aims to integrate
insights from symbolism and connectionism. The paper concludes with
some speculations about extending the present ideas.

Item Type:	Report
Report Nr:	PP-2004-16
Series Name:	Prepublication (PP) Series
Year:	2004
Uncontrolled Keywords:	Subsymbolic Computation, Neural Networks
Depositing User:	Benedikt
Date Deposited:	12 Oct 2016 14:36
Last Modified:	12 Oct 2016 14:36
URI:	https://eprints.illc.uva.nl/id/eprint/127

Actions (login required)

View Item