From Bandits to Experts: A Tale of Domination and Independence
نویسندگان
چکیده
We consider the partial observability model for multi-armed bandits, introduced by Mannor and Shamir [11]. Our main result is a characterization of regret in the directed observability model in terms of the dominating and independence numbers of the observability graph. We also show that in the undirected case, the learner can achieve optimal regret without even accessing the observability graph before selecting an action. Both results are shown using variants of the Exp3 algorithm operating on the observability graph in a time-efficient manner.
منابع مشابه
From Bandits to Experts: A Tale of Domination and Independence
We consider the partial observability model for multi-armed bandits, introducedby Mannor and Shamir [11]. Our main result is a characterization of regret inthe directed observability model in terms of the dominating and independencenumbers of the observability graph. We also show that in the undirected case, thelearner can achieve optimal regret without even accessing the observ...
متن کاملMixed Roman domination and 2-independence in trees
Let $G=(V, E)$ be a simple graph with vertex set $V$ and edge set $E$. A {em mixed Roman dominating function} (MRDF) of $G$ is a function $f:Vcup Erightarrow {0,1,2}$ satisfying the condition that every element $xin Vcup E$ for which $f(x)=0$ is adjacentor incident to at least one element $yin Vcup E$ for which $f(y)=2$. The weight of anMRDF $f$ is $sum _{xin Vcup E} f(x)$. The mi...
متن کاملI.R. and Independence-Seeking Based in the Ideal Sky and Civilized Horizon
In this study the impact of the Islamic revolution on freedom as an important achievement, and the changes and levels of independence in the growing process of the revolution. In a negative approach, the research proposes the hypothesis that the Islamic Revolution has opposed the Westphalian state-based independence and developed Islamic teaching-based independence based. The present study, wit...
متن کاملCoverings, matchings and paired domination in fuzzy graphs using strong arcs
The concepts of covering and matching in fuzzy graphs using strong arcs are introduced and obtained the relationship between them analogous to Gallai’s results in graphs. The notion of paired domination in fuzzy graphs using strong arcs is also studied. The strong paired domination number γspr of complete fuzzy graph and complete bipartite fuzzy graph is determined and obtained bounds for the s...
متن کاملThe Rule of "Nafye Sabil" [i.e. to prevent the Islamic society to be dominated by non-Muslims] in Islamic Thought and Foreign Policy of Islamic Republic of Iran
As a jurisprudential rule, "nafye sabil" has played a sustaining and influential role in Islamic system's major decisions, policies and behavior. This principle is of high importance in Islamic state's foreign relations. Rejecting oppression and tyranny against Muslims and preserving their freedom and removing dependence on aliens is the foundation of this rule in Islamic republic of Iran's for...
متن کامل