Stochastic Contextual Dueling Bandits under Linear Stochastic Transitivity Models
This work addresses the challenge of efficient decision-making in preference-based learning with contextual information, which is incremental as it refines existing models and algorithms.
The paper tackles the problem of minimizing regret in contextual dueling bandits under a linear stochastic transitivity model, proposing an algorithm called CoLSTIM that achieves a regret of order O(√(dT)) after T rounds, with experiments showing superiority over state-of-the-art methods.
We consider the regret minimization task in a dueling bandits problem with context information. In every round of the sequential decision problem, the learner makes a context-dependent selection of two choice alternatives (arms) to be compared with each other and receives feedback in the form of noisy preference information. We assume that the feedback process is determined by a linear stochastic transitivity model with contextualized utilities (CoLST), and the learner's task is to include the best arm (with highest latent context-dependent utility) in the duel. We propose a computationally efficient algorithm, $\texttt{CoLSTIM}$, which makes its choice based on imitating the feedback process using perturbed context-dependent utility estimates of the underlying CoLST model. If each arm is associated with a $d$-dimensional feature vector, we show that $\texttt{CoLSTIM}$ achieves a regret of order $\tilde O( \sqrt{dT})$ after $T$ learning rounds. Additionally, we also establish the optimality of $\texttt{CoLSTIM}$ by showing a lower bound for the weak regret that refines the existing average regret analysis. Our experiments demonstrate its superiority over state-of-art algorithms for special cases of CoLST models.