Source code for Policies.LearnExp

# -*- coding: utf-8 -*-
r""" The LearnExp aggregation bandit algorithm, similar to Exp4 but not equivalent.

The algorithm is a master A, managing several "slave" algorithms, :math:`A_1, ..., A_N`.

- At every step, one slave algorithm is selected, by a random selection from a trust distribution on :math:`[1,...,N]`.
- Then its decision is listen to, played by the master algorithm, and a feedback reward is received.
- The reward is reweighted by the trust of the listened algorithm, and given back to it *with* a certain probability.
- The other slaves, whose decision was not even asked, receive nothing.
- The trust probabilities are first uniform, :math:`P_i = 1/N`, and then at every step, after receiving the feedback for *one* arm k (the reward), the trust in each slave Ai is updated: :math:`P_i` by the reward received.
- The detail about how to increase or decrease the probabilities are specified in the reference article.

.. note:: Reference: [[Learning to Use Learners' Advice, A.Singla, H.Hassani & A.Krause, 2017](https://arxiv.org/abs/1702.04825)].
"""
from __future__ import division, print_function  # Python 2 compatibility

__author__ = "Lilian Besson"
__version__ = "0.7"

from random import random
import numpy as np
import numpy.random as rn
try:
    from .BasePolicy import BasePolicy
    from .with_proba import with_proba
except ImportError:
    from BasePolicy import BasePolicy
    from with_proba import with_proba


# --- Renormalize function

[docs]def renormalize_reward(reward, lower=0., amplitude=1., trust=1., unbiased=True, mintrust=None): r"""Renormalize the reward to `[0, 1]`: - divide by (`trust/mintrust`) if `unbiased` is `True`. - simply project to `[0, 1]` if `unbiased` is `False`, .. warning:: If `mintrust` is unknown, the unbiased estimator CANNOT be projected back to a bounded interval. """ if unbiased: if mintrust is not None: return (reward - lower) / (amplitude * (trust / mintrust)) else: return (reward - lower) / (amplitude * trust) else: return (reward - lower) / amplitude
[docs]def unnormalize_reward(reward, lower=0., amplitude=1.): r"""Project back reward to `[lower, lower + amplitude]`.""" return lower + (reward * amplitude)
# --- Parameters for the LearnExp algorithm # Default values for the parameters #: self.unbiased is a flag to know if the rewards are used as biased estimator, #: i.e., just :math:`r_t`, or unbiased estimators, :math:`r_t / p_t`, if :math:`p_t` is the probability of selecting that arm at time :math:`t`. #: It seemed to work better with unbiased estimators (of course). UNBIASED = False UNBIASED = True # Better #: Default value for the constant Eta in (0, 1] ETA = 0.5 # --- LearnExp algorithm
[docs]class LearnExp(BasePolicy): """ The LearnExp aggregation bandit algorithm, similar to Exp4 but not equivalent."""
[docs] def __init__(self, nbArms, children=None, unbiased=UNBIASED, eta=ETA, prior='uniform', lower=0., amplitude=1. ): # Attributes self.nbArms = nbArms #: Number of arms. self.lower = lower #: Lower values for rewards. self.amplitude = amplitude #: Larger values for rewards. self.unbiased = unbiased #: Flag, see above. assert 0 < eta < 1, "Error: parameter 'eta' for a LearnExp player was expected to be in (0, 1) but = {:.3g}...".format(eta) # DEBUG if eta is None: eta = 1. / nbArms self.eta = eta #: Constant parameter :math:`\eta`. self.nbChildren = nbChildren = len(children) #: Number N of slave algorithms. self.rate = eta / nbChildren #: Constant :math:`\eta / N`, faster computations if it is stored once. assert self.rate > 0, "Error: parameter 'rate' for a LearnExp player was expected to be > 0, but = {:.3g}...".format(self.rate) # DEBUG # Internal object memory self.children = [] #: List of slave algorithms. for i, child in enumerate(children): if isinstance(child, dict): print(" Creating this child player from a dictionnary 'children[{}]' = {} ...".format(i, child)) # DEBUG localparams = {'lower': lower, 'amplitude': amplitude} localparams.update(child['params']) self.children.append(child['archtype'](nbArms, **localparams)) elif isinstance(child, type): print(" Using this not-yet created player 'children[{}]' = {} ...".format(i, child)) # DEBUG self.children.append(child(nbArms, lower=lower, amplitude=amplitude)) # Create it here! else: print(" Using this already created player 'children[{}]' = {} ...".format(i, child)) # DEBUG self.children.append(child) self.last_choice = None #: Remember the index of the last child trusted for a decision. # Initialize the arrays # Assume uniform prior if not given or if = 'uniform' self.trusts = np.full(nbChildren, 1. / nbChildren) #: Initial trusts in the slaves :math:`p_j^t`. Default to uniform, but a prior can also be given. if prior is not None and prior != 'uniform': assert len(prior) == nbChildren, "Error: the 'prior' argument given to LearnExp has to be an array of the good size ({}).".format(nbChildren) # DEBUG self.trusts = prior self.weights = (self.trusts - self.rate) / (1 - self.eta) #: Weights :math:`w_j^t`.
[docs] def __str__(self): """ Nicely print the name of the algorithm with its relevant parameters.""" is_unbiased = "" if self.unbiased else ", biased" return r"LearnExp($N={}${}, $\eta={:.3g}$)".format(self.nbChildren, is_unbiased, self.eta)
# --- Start the game
[docs] def startGame(self): """ Start the game for each child.""" # Start all children for i in range(self.nbChildren): self.children[i].startGame()
# --- Get a reward
[docs] def getReward(self, arm, reward): """ Give reward for each child, and then update the trust probabilities.""" reward = float(reward) new_reward = renormalize_reward(reward, lower=self.lower, amplitude=self.amplitude, unbiased=False) # print(" A LearnExp player {} received a reward = {:.3g} on arm {} and trust = {:.3g} on that choice = {}, giving {:.3g} ...".format(self, reward, arm, self.trusts[self.last_choice], self.last_choice, new_reward)) # DEBUG # 1. First, give rewards to that slave, with probability rate / trusts probability = self.rate / self.trusts[self.last_choice] assert 0 <= probability <= 1, "Error: 'probability' = {:.3g} = rate = {:.3g} / trust_j^t = {:.3g} should have been in [0, 1]...".format(probability, self.rate, self.trusts[self.last_choice]) # DEBUG if with_proba(probability): self.children[self.last_choice].getReward(arm, reward) # 2. Then reinitialize this array of losses assert 0 <= new_reward <= 1, "Error: the normalized reward {:.3g} was NOT in [0, 1] ...".format(new_reward) # DEBUG loss = (1 - new_reward) if self.unbiased: loss /= self.trusts[self.last_choice] # 3. Update weight of that slave self.weights[self.last_choice] *= np.exp(- self.rate * loss) # 4. Recomputed the trusts from the weights # add uniform mixing of proportion rate=eta/N self.trusts = (1 - self.eta) * (self.weights / np.sum(self.weights)) + self.rate # self.trusts = trusts / np.sum(trusts) # XXX maybe this isn't necessary... # print(" The most trusted child policy is the {}th with confidence {}...".format(1 + np.argmax(self.trusts), np.max(self.trusts))) # DEBUG assert np.isclose(np.sum(self.trusts), 1), "Error: 'trusts' do not sum to 1 but to {:.3g} instead...".format(np.sum(self.trusts)) # DEBUG
# print("self.trusts =", self.trusts) # DEBUG # --- Choice of arm methods
[docs] def choice(self): """ Trust one of the slave and listen to his `choice`.""" # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].choice()
[docs] def choiceWithRank(self, rank=1): """ Trust one of the slave and listen to his `choiceWithRank`.""" # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].choiceWithRank(rank=rank)
[docs] def choiceFromSubSet(self, availableArms='all'): """ Trust one of the slave and listen to his `choiceFromSubSet`.""" # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].choiceFromSubSet(availableArms=availableArms)
[docs] def choiceMultiple(self, nb=1): """ Trust one of the slave and listen to his `choiceMultiple`.""" # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].choiceMultiple(nb=nb)
[docs] def choiceIMP(self, nb=1, startWithChoiceMultiple=True): """ Trust one of the slave and listen to his `choiceIMP`.""" # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].choiceIMP(nb=nb)
[docs] def estimatedOrder(self): r""" Trust one of the slave and listen to his `estimatedOrder`. - Return the estimate order of the arms, as a permutation on :math:`[0,...,K-1]` that would order the arms by increasing means. """ # 1. first decide who to listen to self.last_choice = rn.choice(self.nbChildren, p=self.trusts) # 2. then listen to him return self.children[self.last_choice].estimatedOrder()
[docs] def estimatedBestArms(self, M=1): """ Return a (non-necessarily sorted) list of the indexes of the M-best arms. Identify the set M-best.""" assert 1 <= M <= self.nbArms, "Error: the parameter 'M' has to be between 1 and K = {}, but it was {} ...".format(self.nbArms, M) # DEBUG order = self.estimatedOrder() return order[-M:]