Extended encoding/decoding of embedded structures using connectionist networks

1999 ◽  
Author(s):  
L. Niklasson
Keyword(s):  
Connectionist Networks ◽  
Embedded Structures

Recursive generation in language

2017 ◽  
Author(s):  
David J. Lobina
Keyword(s):  
Mathematical Logic ◽  
Production System ◽  
General Property ◽  
Recent Literature ◽  
Expressive Power ◽  
The Self ◽  
Recursive Structure ◽  
Embedded Structures ◽  
Mechanical Procedure ◽  
The 1950S

The introduction of recursion into linguistics was the result of applying some of the results of mathematical logic to the study of language. In particular, recursion was introduced in the 1950s as a general property of the mechanical procedure underlying the grammar, in order to account for language’s discrete infinity and expressive power—in the 1950s, this mechanical procedure was a production system, whereas more recently, of course, it is the set-operator merge. Unfortunately, the recent literature has confused the general recursive property of a grammar with specific instances of (recursive) rules/operations within a grammar; more worryingly still, there has been a general conflation of these recursive rules with some of the self-embedded structures these rules can generate, adding to the confusion. The conflation is manifold but always fallacious. Moreover, language manifests a much more generally recursive structure than is usually recognized: bundles of the universal (Specifier)-Head-Complement(s) geometry.

Using Noise to Compute Error Surfaces in Connectionist Networks: A Novel Means of Reducing Catastrophic Forgetting

2002 ◽  
Vol 14 (7) ◽  
pp. 1755-1769 ◽  
Author(s):  
Robert M. French ◽  
Nick Chater
Keyword(s):  
Feedforward Networks ◽  
First Derivative ◽  
Error Surface ◽  
Connectionist Networks ◽  
New Information ◽  
Catastrophic Interference ◽  
Derivative Information

In error-driven distributed feedforward networks, new information typically interferes, sometimes severely, with previously learned information. We show how noise can be used to approximate the error surface of previously learned information. By combining this approximated error surface with the error surface associated with the new information to be learned, the network's retention of previously learned items can be improved and catastrophic interference significantly reduced. Further, we show that the noise-generated error surface is produced using only first-derivative information and without recourse to any explicit error information.

Vibrations of Foundations Interacting With Subsoil Experimental Studies and BE Calculations

1991 ◽  
Author(s):  
L. Gaul
Keyword(s):  
Boundary Element Method ◽  
Dynamic Response ◽  
Surface Waves ◽  
Boundary Element ◽  
Constitutive Equations ◽  
Experimental Studies ◽  
Boundary Elements ◽  
Experimental Techniques ◽  
Embedded Structures ◽  
Element Method

Abstract Calculation of the dynamic response of sensitive structures like foundations for vibrating machinery requires to take the interaction with subsoil into account. Structures and soil are discretized by boundary elements and coupled by a substructure technique. Viscoelastic constitutive equations contain fractional time derivatives. Surface waves generated by machine foundations and diffracted by embedded structures and soil inhomogeneities are analyzed by conventional and optoelectronic experimental techniques and calculated by the boundary element method (BEM).

Sign in / Sign up

Export Citation Format

Share Document