d3rlpy.online.explorers.LinearDecayEpsilonGreedy

class d3rlpy.online.explorers.LinearDecayEpsilonGreedy(start_epsilon=1.0, end_epsilon=0.1, duration=1000000)

\(\epsilon\)-greedy explorer with linear decay schedule.

Parameters:

• start_epsilon (float) – the beginning \(\epsilon\).
• end_epsilon (float) – the end \(\epsilon\).
• duration (int) – the scheduling duration.

start_epsilon

the beginning \(\epsilon\).

Type:float

end_epsilon

the end \(\epsilon\).

Type:float

duration

the scheduling duration.

Type:int

Methods

compute_epsilon(step)

Returns decayed \(\epsilon\).

Returns:\(\epsilon\). Return type:float

sample(algo, x, step)

Returns \(\epsilon\)-greedy action.

Parameters:

• algo (d3rlpy.algos.base.AlgoBase) – algorithm.
• x (numpy.ndarray) – observation.
• step (int) – current environment step.

Returns:

\(\epsilon\)-greedy action.

Return type:

int