Reinforcement Learning — Mini Glossary — EN/TRAgent (Ajan): An agent acquires decision-making skills through trial and error, guided by rewards and punishments from its surroundings.Jun 9, 2023Jun 9, 2023
The “Deep” in Reinforcement LearningDeep Reinforcement Learning incorporates deep neural networks into the framework of Reinforcement Learning. This integration allows for…May 31, 2023May 31, 2023
Two main approaches for solving RL problems: Policy-Based Methods/Value-Based MethodsPolicy-Based MethodsMay 31, 2023May 31, 2023
The Exploration/Exploitation trade-offThe exploration/exploitation trade-off is a fundamental concept in reinforcement learning that refers to the dilemma of choosing between…May 31, 2023May 31, 2023
Identifying reward functions and the concept of discounted rewardsIn reinforcement learning (RL), the reward serves as the fundamental feedback for the agent’s actions.May 31, 2023May 31, 2023
Observations/States SpaceObservations: Observations refer to the information that an agent receives from the environment. In the context of reinforcement learning…May 31, 2023May 31, 2023
Markov PropertyThe Markov Property in Markov Decision Processes (MDPs) is a fundamental concept that significantly impacts the agent’s decision-making…May 31, 2023May 31, 2023
The reward hypothesisIn reinforcement learning, the learning process typically follows a loop that generates a sequence of state-action-reward-next state…May 31, 2023May 31, 2023
How does Reinforcement Learning work?The agent receives the initial state, denoted as S₀, from the environment. In this case, the state represents the first frame of a game.May 31, 2023May 31, 2023
