By Topic

Design of neural network quantizers for a distributed estimation system with communication constraints

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Megalooikonomou, V. ; Dept. of Comput. Sci. & Electr. Eng., Maryland Univ., Baltimore, MD, USA ; Yesha, Y.

We consider the problem of quantizer design in a distributed estimation system with communication constraints at the channels in the case where the observation statistics are unknown and one must rely on a training set. The method that we propose applies a variation of the cyclic generalized Lloyd algorithm (CGLA) on every point of the training set and then uses a neural network for each quantizer to represent the training points along with their associated codewords. The codeword of every training point is initialized using a regression tree approach. Simulations show that the combined approach i.e. building the regression tree system and using its quantizers to initialize the neural networks provides an improvement over the regression tree approach except in the case of high noise variance

Published in:

Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on  (Volume:6 )

Date of Conference:

12-15 May 1998