Minimising Contrastive Divergence in Noisy, Mixed-mode VLSI Neurons

This paper presents VLSI circuits with continuous-valued probabilistic behaviour realized by injecting noise into each computing unit(neuron). Interconnecting the noisy neurons forms a Continuous Restricted Boltzmann Machine (CRBM), which has shown promising performance in modelling and classifying noisy biomedical data. The Minimising-Contrastive-Divergence learning algorithm for CRBM is also implemented in mixed-mode VLSI, to adapt the noisy neurons’ parameters on-chip.

• Type: Conference
• Year: 2003
• Where: NIPS
• Authors: Hsin Chen, Patrice Fleury, Alan F. Murray
• Keywords: NIPS 2003, NIPS 2007, Noisy Biomedical Data, Noisy Neurons, Paper Presents Vlsi