The dueling bandit is a learning framework wherein the feedback information in the learning process is restricted to a noisy comparison between a pair of actions. In this research, we address a dueling bandit problem based on a cost function over a continuous space. We propose a stochastic mirror descent algorithm and show that the algorithm achieves an O( √ T log T )-regret bound under strong ...

BACKGROUND A tube feeding decision aid designed at the Ottawa Health Research Institute was specifically created for substitute decision-makers who must decide whether to allow placement of a percutaneous endoscopic gastrostomy (PEG) tube in a cognitively impaired older person. We developed a Japanese version and found that the decision aid promoted the decision-making process of substitute dec...

We describe and analyze an algorithmic framework for playing convex repeatedgames. In each trial of the repeated game, the first player predicts a vector andthen the second player responds with a loss function over the vector. Based on ageneralization of Fenchel duality, we derive an algorithmic framework for the firstplayer and analyze the player’s regret. We then use our a...

Sensitive error correcting output codes are a reduction from cost sensitive classi cation to binary classi cation. They are a modi cation of error correcting output codes [3] which satisfy an additional property: regret for binary classi cation implies at most 2 l2 regret for cost-estimation. This has several implications: 1) Any 0/1 regret minimizing online algorithm is (via the reduction) a r...

We study the problem of online kernel selection under computational constraints, where memory or time and prediction procedures is restricted to a fixed budget. In this paper, we analyze worst-case lower bounds on regret algorithm with subset observed examples, design algorithms enjoying corresponding upper bounds. also identify condition which constraints different from that constraints. To al...

A general class of no-regret learning algorithms, called no-Φ-regret learning algorithms, is defined which spans the spectrum from no-external-regret learning to no-internal-regret learning and beyond. The set Φ describes the set of strategies to which the play of a given learning algorithm is compared. A learning algorithm satisfies no-Φ-regret if no regret is experienced for playing as the al...

We consider online content recommendation with implicit feedback through pairwise comparisons, formalized as the so-called dueling bandit problem. We study the dueling bandit problem in the Condorcet winner setting, and consider two notions of regret: the more well-studied strong regret, which is 0 only when both arms pulled are the Condorcet winner; and the less well-studied weak regret, which...