RLAutoMLAgentType
Enums · AiDotNet.Enums
Defines which reinforcement learning agent families can be explored by AutoML.
For Beginners
Different RL agents are better suited for different problems:
DQNis popular for discrete action spaces (like left/right).PPOis a strong general-purpose agent (discrete or continuous).A2Cis a simple actor-critic baseline.DDPGandSACare commonly used for continuous control.
How It Works
This enum is used by facade configuration options to select which RL agent types AutoML is allowed to try.
Fields
| Field | Summary |
|---|---|
RLAutoMLAgentType A2C |
Advantage Actor-Critic (A2C) for discrete or continuous control. |
RLAutoMLAgentType DDPG |
Deep Deterministic Policy Gradient (DDPG) for continuous control. |
RLAutoMLAgentType DQN |
Deep Q-Network (DQN) for discrete action spaces. |
RLAutoMLAgentType PPO |
Proximal Policy Optimization (PPO) for discrete or continuous control. |
RLAutoMLAgentType SAC |
Soft Actor-Critic (SAC) for continuous control. |
int value__ |