arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1306.3895cs.LG
收藏

具有最优遗憾的在线PCA

On-line PCA with Optimal Regrets

University of California, Santa Cruz(加州大学圣克鲁兹分校);Poznań University of Technology(波兹南理工大学)

Jiazhong Nie, Wojciech Kotlowski, Manfred K. Warmuth

摘要:

我们仔细研究了PCA的在线版本,其中在每次试验中,一个学习算法扮演一个k维子空间,并且在投影到所选择的子空间时,在下一个实例上遭受压缩损失。在这种情况下,我们分析了两个流行的在线算法,梯度下降(GD)和指数梯度(EG)。我们表明,这两种算法在最坏的情况下基本上是最优的。这是一个惊喜,因为EG是已知的执行次优时,实例稀疏。EG对于PCA的这种不同行为主要与这种情况下损失的非负性有关,这使得PCA设置与文献中研究的其他设置有质的不同。此外,我们表明,当考虑遗憾界作为损失预算的函数,EG仍然是最优的,严格优于GD。接下来,我们研究PCA设置的扩展,其中自然被允许使用稠密实例,这些实例是具有有界最大特征值的正矩阵。我们可以再次证明,在这种情况下,EG是最佳的,并且严格优于GD。

英文摘要:

We carefully investigate the on-line version of PCA, where in each trial a learning algorithm plays a k-dimensional subspace, and suffers the compression loss on the next instance when projected into the chosen subspace. In this setting, we analyze two popular on-line algorithms, Gradient Descent (GD) and Exponentiated Gradient (EG). We show that both algorithms are essentially optimal in the worst-case. This comes as a surprise, since EG is known to perform sub-optimally when the instances are sparse. This different behavior of EG for PCA is mainly related to the non-negativity of the loss in this case, which makes the PCA setting qualitatively different from other settings studied in the literature. Furthermore, we show that when considering regret bounds as function of a loss budget, EG remains optimal and strictly outperforms GD. Next, we study the extension of the PCA setting, in which the Nature is allowed to play with dense instances, which are positive matrices with bounded largest eigenvalue. Again we can show that EG is optimal and strictly better than GD in this setting.

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。

↑