Glossary¶
- HPO¶
Hyperparameter Optimization. The process of finding the best hyperparameters for a given machine learning model.
- AutoRL¶
Automated Reinforcement Learning. The process of automating the process of applying reinforcement learning to real-world problems.
- AutoML¶
Automated Machine Learning. The process of automating the process of applying machine learning to real-world problems.
- Random Search¶
A hyperparameter optimization algorithm that randomly samples hyperparameters from a predefined search space.
- BO¶
Bayesian Optimization. A Black-Box optimization algorithm weighing exploration & exploitation to find the minimum of its objective.
- Multi-fidelity optimization¶
The process of optimizing a function using multiple levels of fidelity. This can be done by using a surrogate model to approximate the function at different levels of fidelity.
- PPO¶
Proximal Policy Optimization. A policy gradient method that uses a clipped surrogate objective to improve the stability of the learning process.
- DQN¶
Deep Q-Network. A deep reinforcement learning algorithm that uses a neural network to approximate the Q-function.
- SAC¶
Soft Actor-Critic. An off-policy reinforcement learning algorithm that uses the maximum entropy principle to improve the exploration of the environment.