Rapid competitive learning techniques for neural networks

US10846595B2 · kind B2 · utility

Assignee

• Intel Corporation · US

Inventors

• Andreas Wild · Weimar, DE
• Narayan Srinivasa · Thousand Oaks, US

Key dates

• Filing date: Dec 20, 2016
• Grant date: Nov 24, 2020
• Expiry date: Sep 25, 2039

Classification

• Technology area (CPC G): Physics
• CPC primary: G06N3/049
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

Various systems and methods for implementing unsupervised or reinforcement learning operations for a neuron weight used in a neural network are described. In an example, the learning operations include processing a spike train input at a neuron of a spiking neural network, applying a synaptic weight, and observing spike events occurring before and after the neuron processing based on respective spike traces. A synaptic weight update process operates to generate a new value of the synaptic weight based upon the spike traces, configuration values, and a reference weight value. A reference weight update process also operates to generate a new value of the reference value for significant changes to the synaptic weight. Reinforcement may be provided in some examples to implement changes to the reference weight in reduced time. In some examples, the techniques may be implemented in a neuromorphic hardware implementation of the spiking neural network.