Conceptlocal learning rules
Local Learning Rules
A class of machine learning methods in which each computational unit updates its own parameters using only locally available information, without requiring global gradient propagation. This approach naturally supports online learning scenarios. Representative methods include Hebbian learning and competitive learning.
Timeline (last 90 days)
Sep 12
In local learning rules, each computational unit updates itself based only on locally available information without relying on global gradient back-propagation, making it naturally suited for online learning scenarios.
Unverified50%