Eligibility Traces
Eligibility traces are a temporary memory mechanism in reinforcement learning that helps assign credit to recent actions for future rewards.
106 plain-language definitions from the TiorAI glossary, filed under Reinforcement Learning. Every entry opens with a one-sentence definition, then explains where the term is used.
Eligibility traces are a temporary memory mechanism in reinforcement learning that helps assign credit to recent actions for future rewards.
An episode is a single installment or segment within a larger series, typically in television, podcasts, or web content.
Epsilon-Greedy is a strategy used in decision-making algorithms to balance exploration and exploitation by choosing random actions with a small probability epsilon and the best-known action otherwise.
Evolutionary Game Theory is a framework that applies game theory principles to evolving populations in biology and social sciences, analyzing how strategies evolve over time through natural selection or imitation.
Expansion is the process of growing or increasing the size, scope, or reach of a business, market, or concept.
Expected SARSA is a reinforcement learning algorithm that updates action-value estimates by considering the expected value of all possible next actions under the current policy.
Exploration-Exploitation Tradeoff is the decision-making dilemma between exploring new options and exploiting known resources to maximize rewards.