Also found in: Medical, Wikipedia.


References in periodicals archive ?
Denzinger & Laureyns, 2008) have concluded that SARSA algorithm (Sutton & Barto, 1998) is a possible solution in the motion control of a 2-DOF manipulator system.
The standard SARSA algorithm is applied with various reinforcement learning parameters of [alpha] = 0.
Using a variety of reinforcement learning algorithms (Q-learning, SARSA, and prioritized sweeping) we experimented with a simple 8 by 8 grid world with rewards in cells (1, 1) and (8, 8).