Repository logo

Infoscience

  • English
  • French
Log In
Logo EPFL, École polytechnique fédérale de Lausanne

Infoscience

  • English
  • French
Log In
  1. Home
  2. Academic and Research Output
  3. Conferences, Workshops, Symposiums, and Seminars
  4. Connectionist Quantization Functions
 
conference paper

Connectionist Quantization Functions

Lundin, Tomas
•
Fiesler, Emile
•
Moerland, Perry
1996
Proceedings of the '96 SIPAR-Workshop on Parallel and Distributed Computing
Scientific and Parallel Computing Group, University of Geneva - Proceedings of the '96 SIPAR-Workshop on Parallel and Distributed Computing

One of the main strengths of connectionist systems, also known as neural networks, is their massive parallelism. However, most neural networks are simulated on serial computers where the advantage of massive parallelism is lost. For large and real-world applications, parallel hardware implementations are therefore essential. Since a discretization or quantization of the neural network parameters is of great benefit for both analog and digital hardware implementations, they are the focus of study in this paper. In 1987 a successful weight discretization method was developed, which is flexible and produces networks with few discretization levels and without significant loss of performance. However, recent studies have shown that the chosen quantization function is not optimal. In this paper, new quantization functions are introduced and evaluated for improving the performance of this flexible weight discretization method.

  • Files
  • Details
  • Metrics
Loading...
Thumbnail Image
Name

96sipar.pdf

Access type

openaccess

Size

154.05 KB

Format

Adobe PDF

Checksum (MD5)

60ccc30b0d7c4b4c9585c6951fbb0eef

Logo EPFL, École polytechnique fédérale de Lausanne
  • Contact
  • infoscience@epfl.ch

  • Follow us on Facebook
  • Follow us on Instagram
  • Follow us on LinkedIn
  • Follow us on X
  • Follow us on Youtube
AccessibilityLegal noticePrivacy policyCookie settingsEnd User AgreementGet helpFeedback

Infoscience is a service managed and provided by the Library and IT Services of EPFL. © EPFL, tous droits réservés