作者: Filip Piekniewski
DOI:
关键词:
摘要: Apparatus and methods for conditional plasticity in a neural network. In one approach, mechanism is configured to select alternate rules when performing connection updates. The selection adapted based on comparison of actual efficiency target efficiency. For instance, below the value, STDP rule may be modulated increase long term potentiation. Similarly, above depression. dynamically adjusts efficacy, prevents uncontrolled weights, thereby improving network operation processing information varying nature.