New method corrects least-squares temporal difference for better lambda-return estimation.
problem Improving lambda-return estimation in reinforcement learning.
method Uncorrected least-squares temporal difference with a correction method.
result Enhanced accuracy in temporal difference learning.