We study the problem of online multiclass classification in a setting where the learner's feedback is determined by an arbitrary directed graph. While including bandit feedback as a special case, feedback graphs allow a much richer set of applications, including filtering and label efficient classification. We introduce Gappletron, the first online multiclass algorithm that works with arbitrary feedback graphs. For this new algorithm, we prove surrogate regret bounds that hold, both in expectation and with high probability, for a large class of surrogate losses. Our bounds are of order $B\sqrt{\rho KT}$, where $B$ is the diameter of the prediction space, $K$ is the number of classes, $T$ is the time horizon, and $\rho$ is the domination number (a graph-theoretic parameter affecting the amount of exploration). In the full information case, we show that Gappletron achieves a constant surrogate regret of order $B^2K$. We also prove a general lower bound of order $\max\big\{B^2K,\sqrt{T}\big\}$ showing that our upper bounds are not significantly improvable. Experiments on synthetic data show that for various feedback graphs, our algorithm is competitive against known baselines.
翻译:我们研究一个环境中的在线多级分类问题,在这个环境中,学习者的反馈是由任意的定向图表决定的。在将土匪反馈作为特例的情况下,反馈图表允许更丰富的一系列应用,包括过滤和标签效率分类。我们引入了首个使用任意反馈图表的在线多级算法Gaplertron。对于这个新的算法,我们证明代孕后悔恨,这种悔恨在预期和可能性很大的情况下,维持着一大批代孕损失。我们的底线是$B\sqrt=rho KT}美元,其中美元是预测空间直径,美元是类别数,T$是时间范围,而$rho$是支配数(一个影响探索量的图形理论参数)。在全部信息案例中,我们证明Gappletron的常年代代代代孕後悔是 $B2K美元。我们还证明,对于预测空间直径为$B&B2K, $B2K, $K, $K, $K, $K, $, $, $T\\bigrt\\\\\\\\ 美元, 美元是分类数数数数, 值是用来显示我们已知的顶端数据。