Approximate information maximization for bandit games
Résumé
Entropy maximization and free energy minimization are general physical principles for modeling the dynamics of various physical systems. Notable examples include modeling decision-making within the brain using the free-energy principle, optimizing the accuracy-complexity trade-off when accessing hidden variables with the information bottleneck principle (Tishby et al., 2000), and navigation in random environments using information maximization (Vergassola et al., 2007). Built on this principle, we propose a new class of bandit algorithms that maximize an approximation to the information of a key variable within the system. To this end, we develop an approximated analytical physics-based representation of an entropy to forecast the information gain of each action and greedily choose the one with the largest information gain. This method yields strong performances in classical bandit settings. Motivated by its empirical success, we prove its asymptotic optimality for the two-armed bandit problem with Gaussian rewards. Owing to its ability to encompass the system’s properties in a global physical functional, this approach can be efficiently adapted to more complex bandit settings, calling for further investigation of information maximization approaches for multi-armed bandit problems.
Fichier principal
Approximate information maximization for bandit games.pdf (777.89 Ko)
Télécharger le fichier
Figure1.pdf (299.37 Ko)
Télécharger le fichier
Figure2.pdf (206.65 Ko)
Télécharger le fichier
Figure_bernoulli_Aistats.pdf (206.56 Ko)
Télécharger le fichier
Figure_bernoulli_Aistats_close_arms.pdf (206.14 Ko)
Télécharger le fichier
Figure_close_arms.pdf (205.9 Ko)
Télécharger le fichier
Figure_multiple_arms.pdf (206 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)