Labeling RAAM

In this paper we propose an extension of the Recursive Auto-Associative Memory (RAAM) by Pollack. This extension, the Labeling RAAM (LRAAM), is able to encode labeled graphs with cycles by representing pointers explicitly. A theoretical analysis of the constraints imposed on the weights by the learn...

Full description

Bibliographic Details
Main Author: Alessandro Sperduti
Other Authors: The Pennsylvania State University CiteSeerX Archives
Format: Text
Language:English
Published: 1994
Subjects:
Online Access:http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.49.5737
id ftciteseerx:oai:CiteSeerX.psu:10.1.1.49.5737
record_format openpolar
spelling ftciteseerx:oai:CiteSeerX.psu:10.1.1.49.5737 2023-05-15T18:32:40+02:00 Labeling RAAM Alessandro Sperduti The Pennsylvania State University CiteSeerX Archives 1994 application/postscript http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.49.5737 en eng http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.49.5737 Metadata may be used without restrictions as long as the oai identifier remains attached to it. ftp://ftp.icsi.berkeley.edu/pub/techreports/1993/tr-93-029.ps.gz text 1994 ftciteseerx 2021-06-27T00:20:10Z In this paper we propose an extension of the Recursive Auto-Associative Memory (RAAM) by Pollack. This extension, the Labeling RAAM (LRAAM), is able to encode labeled graphs with cycles by representing pointers explicitly. A theoretical analysis of the constraints imposed on the weights by the learning task under the hypothesis of perfect learning and linear output units is presented. Cycles and confluent pointers result to be particularly effective in imposing constraints on the weights. Some technical problems encountered in the RAAM, such as the termination problem in the learning and decoding processes, are solved more naturally in the LRAAM framework. The representations developed for the pointers seem to be robust to recurrent decoding along a cycle. Data encoded in a LRAAM can be accessed by pointer as well as by content. The direct access by content can be achieved by transforming the encoder network of the LRAAM in a Bidirectional Associative Memory (BAM). Different access pro. Text The Pointers Unknown
institution Open Polar
collection Unknown
op_collection_id ftciteseerx
language English
description In this paper we propose an extension of the Recursive Auto-Associative Memory (RAAM) by Pollack. This extension, the Labeling RAAM (LRAAM), is able to encode labeled graphs with cycles by representing pointers explicitly. A theoretical analysis of the constraints imposed on the weights by the learning task under the hypothesis of perfect learning and linear output units is presented. Cycles and confluent pointers result to be particularly effective in imposing constraints on the weights. Some technical problems encountered in the RAAM, such as the termination problem in the learning and decoding processes, are solved more naturally in the LRAAM framework. The representations developed for the pointers seem to be robust to recurrent decoding along a cycle. Data encoded in a LRAAM can be accessed by pointer as well as by content. The direct access by content can be achieved by transforming the encoder network of the LRAAM in a Bidirectional Associative Memory (BAM). Different access pro.
author2 The Pennsylvania State University CiteSeerX Archives
format Text
author Alessandro Sperduti
spellingShingle Alessandro Sperduti
Labeling RAAM
author_facet Alessandro Sperduti
author_sort Alessandro Sperduti
title Labeling RAAM
title_short Labeling RAAM
title_full Labeling RAAM
title_fullStr Labeling RAAM
title_full_unstemmed Labeling RAAM
title_sort labeling raam
publishDate 1994
url http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.49.5737
genre The Pointers
genre_facet The Pointers
op_source ftp://ftp.icsi.berkeley.edu/pub/techreports/1993/tr-93-029.ps.gz
op_relation http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.49.5737
op_rights Metadata may be used without restrictions as long as the oai identifier remains attached to it.
_version_ 1766216864847888384