banditpylib.learners.mab_fcbai_learner

Classes

• MABFixedConfidenceBAILearner: Base class for bai learners in the ordinary multi-armed bandit

• ExpGap: Exponential-gap elimination policy [KKS13]

• LilUCBHeuristic: LilUCB heuristic policy [JMNB14]

class banditpylib.learners.mab_fcbai_learner.MABFixedConfidenceBAILearner(arm_num: int, confidence: float, name: Optional[str])

Base class for bai learners in the ordinary multi-armed bandit

This learner aims to identify the best arm with fixed confidence.

Parameters

• arm_num (int) – number of arms

• confidence (float) – confidence level. It should be within (0, 1). The algorithm should output the best arm with probability at least this value.

• name (str) – alias name

Inheritance

property arm_num: int

Number of arms

abstract property best_arm: int

Index of the best arm identified by the learner

property confidence: float

Confidence level of the learner

property goal: banditpylib.learners.utils.Goal

Goal of the learner

property running_environment: Union[type, List[type]]

Type of bandit environment the learner plays with

class banditpylib.learners.mab_fcbai_learner.ExpGap(arm_num: int, confidence: float, threshold: int = 2, name: Optional[str] = None)

Exponential-gap elimination policy [KKS13]

Parameters

• arm_num (int) – number of arms

• confidence confidence (float) – confidence level. It should be within (0, 1). The algorithm should output the best arm with probability at least this value.

• threshold (int) – do uniform sampling when the active arms are no greater than the threshold within median elimination

• name (Optional[str]) – alias name

Inheritance

actions(context: data_pb2.Context) → data_pb2.Actions

Actions of the learner

Parameters

context – contextual information about the bandit environment

Returns

actions to take

property best_arm: int

Index of the best arm identified by the learner

reset()

Reset the learner

Warning

This function should be called before the start of the game.

property stage: str

Stage of the learner

update(feedback: data_pb2.Feedback)

Update the learner

Parameters

feedback – feedback returned by the bandit environment after actions() is executed

class banditpylib.learners.mab_fcbai_learner.LilUCBHeuristic(arm_num: int, confidence: float, name: Optional[str] = None)

LilUCB heuristic policy [JMNB14]

Parameters

• arm_num (int) – number of arms

• confidence (float) – confidence level. It should be within (0, 1). The algorithm should output the best arm with probability at least this value.

• name (Optional[str]) – alias name

Inheritance

actions(context: data_pb2.Context) → data_pb2.Actions

Actions of the learner

Parameters

context – contextual information about the bandit environment

Returns

actions to take

property best_arm: int

Index of the best arm identified by the learner

reset()

Reset the learner

Warning

This function should be called before the start of the game.

update(feedback: data_pb2.Feedback)

Update the learner

Parameters

feedback – feedback returned by the bandit environment after actions() is executed