UNIGE document Scientific Article
previous document  unige:120889  next document
add to browser collection
Title

Estimating probabilities for unbounded categorization problems

Author
Published in Neurocomputing. 2004, vol. 57, p. 77-86
Abstract We propose two output activation functions for estimating probability distributions over an unbounded number of categories with a recurrent neural network, and derive the statistical assumptions which they embody. Both these methods perform better than the standard approach to such problems, when applied to probabilistic parsing of natural language with Simple Synchrony Networks.
Keywords Recurrent neural networksProbability estimationNatural language parsing
Identifiers
Full text
Structures
Research groups Geneva Artificial Intelligence Laboratory
Laboratoire d'Analyse et de Traitement du Langage (LATL)
Citation
(ISO format)
HENDERSON, James. Estimating probabilities for unbounded categorization problems. In: Neurocomputing, 2004, vol. 57, p. 77-86. doi: 10.1016/j.neucom.2004.01.004 https://archive-ouverte.unige.ch/unige:120889

161 hits

81 downloads

Update

Deposited on : 2019-07-17

Export document
Format :
Citation style :