Glossary term
Glossary term
Agentic Systems
In Deep Q-learning, a neural network that is a stable approximation of the main neural network, where the main neural network implements either a Q-function or a policy. Then, you can train the main network on the Q-values predicted by the target network. Therefore, you prevent the feedback loop that occurs when the main network trains on Q-values predicted by itself. By avoiding this feedback, training stability increases.
Created for this library
A research team uses a target network in its DQN training to stabilize learning over many updates.
A robotics RL team uses a target network in its DQN-based manipulation policy to keep value updates stable.
An energy company uses a target network in its DQN-based HVAC agent to prevent oscillations in value estimates during training.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License