冻结和睡眠:跟踪通过进化过去后验子学习的专家
Freezing and Sleeping: Tracking Experts that Learn by Evolving Past Posteriors
摘要:
弗罗因德提出的一个问题是如何有效地跟踪一个更大的专家集的一小部分。这个问题在Bousquet和Warfall于2001年推出混合过去后验(MPP)算法时得到了解决。在弗罗因德问题中,专家通常被认为是黑箱。然而,在本文中,我们重新审视弗氏问题的情况下,专家有内部结构,使他们能够学习。在这种情况下,问题有两种可能的解释:专家应该从所有数据中学习,还是只从他们正在跟踪的子序列中学习?MPP算法解决了第一种情况。我们的贡献是推广MPP以解决第二种选择。我们得到的结果适用于任何专家结构,可以使用(专家)隐马尔可夫模型进行形式化。奇怪的是,对于我们的解释,有\emph{two}自然的参考方案:冻结和睡眠。对于每个方案,我们提供了一个有效的预测策略,并证明了相关的损失界。
英文摘要:
A problem posed by Freund is how to efficiently track a small pool of experts out of a much larger set. This problem was solved when Bousquet and Warmuth introduced their mixing past posteriors (MPP) algorithm in 2001. In Freund's problem the experts would normally be considered black boxes. However, in this paper we re-examine Freund's problem in case the experts have internal structure that enables them to learn. In this case the problem has two possible interpretations: should the experts learn from all data or only from the subsequence on which they are being tracked? The MPP algorithm solves the first case. Our contribution is to generalise MPP to address the second option. The results we obtain apply to any expert structure that can be formalised using (expert) hidden Markov models. Curiously enough, for our interpretation there are \emph{two} natural reference schemes: freezing and sleeping. For each scheme, we provide an efficient prediction strategy and prove the relevant loss bound.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。