Parallel reinforcement learning with linear function approximation

45 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Asynchronous Methods for Deep Reinforcement Learning (2016, background): “(Grounds & Kudenko, 2008) proposed a parallel version of the Sarsa algorithm that uses multiple separate actor-learners to accelerate training.”
Massively Parallel Methods for Deep Reinforcement Learning (2015, background): “Each machine has its own instance of the agent and environment (Grounds & Kudenko, 2008), running a simple reinforcement learning algorithm (linear Sarsa, in this case).”
Concurrent Reinforcement Learning from Customer Interactions (2013, methodology): “Each time-step has a duration of d; this is typically short and real interactions with each customer are typically sparse.1 We assume that there are a total of N customers; if customers arrive in or exit the system, then their periods of inactivity are represented by null actions.”
Sharing Experience in Multitask Reinforcement Learning (2019, methodology): “In traditional RL approaches [Sutton and Barto, 1998], such as Q-learning and SARSA, linear approximate functions are used to solve the generalization of machine learning algorithms in infinite and dynamic scenarios [Grounds and Kudenko, 2008; Konidaris et al., 2011].”