Upper Confidence Trees and Billiards for Optimal Active Learning
نویسندگان
چکیده
This paper focuses on Active Learning (AL) with bounded computational resources. AL is formalized as a finite horizon Reinforcement Learning problem, and tackled as a single-player game. An approximate optimal AL strategy based on tree-structured multi-armed bandit algorithms and billiard-based sampling is presented together with a proof of principle of the approach. Motsclés : Apprentissage actif, Fouille d’arbre Monte-Carlo, Bandits
منابع مشابه
Confidence Decision Trees via Online and Active Learning for Streaming (BIG) Data
Decision tree classifiers are a widely used tool in data stream mining. The use of confidence intervals to estimate the gain associated with each split leads to very effective methods, like the popular Hoeffding tree algorithm. From a statistical viewpoint, the analysis of decision tree classifiers in a streaming setting requires knowing when enough new information has been collected to justify...
متن کاملThe second geometric-arithmetic index for trees and unicyclic graphs
Let $G$ be a finite and simple graph with edge set $E(G)$. The second geometric-arithmetic index is defined as $GA_2(G)=sum_{uvin E(G)}frac{2sqrt{n_un_v}}{n_u+n_v}$, where $n_u$ denotes the number of vertices in $G$ lying closer to $u$ than to $v$. In this paper we find a sharp upper bound for $GA_2(T)$, where $T$ is tree, in terms of the order and maximum degree o...
متن کاملAn Upper Bound on the First Zagreb Index in Trees
In this paper we give sharp upper bounds on the Zagreb indices and characterize all trees achieving equality in these bounds. Also, we give lower bound on first Zagreb coindex of trees.
متن کاملConfidence Decision Trees via Online and Active Learning for Streaming Data
Decision tree classifiers are a widely used tool in data stream mining. The use of confidence intervals to estimate the gain associated with each split leads to very effective methods, like the popular Hoeffding tree algorithm. From a statistical viewpoint, the analysis of decision tree classifiers in a streaming setting requires knowing when enough new information has been collected to justify...
متن کاملActive Learning on Trees and Graphs
We investigate the problem of active learning on a given tree whose nodes are assigned binary labels in an adversarial way. Inspired by recent results by Guillory and Bilmes, we characterize (up to constant factors) the optimal placement of queries so to minimize the mistakes made on the non-queried nodes. Our query selection algorithm is extremely efficient, and the optimal number of mistakes ...
متن کامل