Deep Q Networks (DQN)
- using deep neural networks to learn Q-function (i.e. Q-learning)
DQN - Architecture Types
---cognitive-computing---machine-intelligence/ai---subfields/machine-learning-(ml)---pattern-recognition/ml---models/reinforcement-learning-(rl)/deep-q-networks-(dqn)/deep-q-network-architecture-types.png)
DQN - Q-Learning (Value Learning vs Policy Learning)
---cognitive-computing---machine-intelligence/ai---subfields/machine-learning-(ml)---pattern-recognition/ml---models/reinforcement-learning-(rl)/deep-q-networks-(dqn)/dqn-value-learning.png)
---cognitive-computing---machine-intelligence/ai---subfields/machine-learning-(ml)---pattern-recognition/ml---models/reinforcement-learning-(rl)/deep-q-networks-(dqn)/dqn-policy-learning.png)
DQN - Continuous Action Space
Policy Learning enables modeling of continuous action space
---cognitive-computing---machine-intelligence/ai---subfields/machine-learning-(ml)---pattern-recognition/ml---models/reinforcement-learning-(rl)/deep-q-networks-(dqn)/dqn-policy-learning-continuous-action-space.png)