arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1406.3140stat.MLmath.PR
收藏

限制玻尔兹曼机的表示能力和逼近误差

Expressive Power and Approximation Errors of Restricted Boltzmann Machines

Guido Montufar, Johannes Rauh, Nihat Ay

摘要:

我们提出了明确的类的概率分布,可以通过限制玻尔兹曼机(RBM)的学习,取决于它们包含的单位的数量,这是代表模型的表达能力。我们用这个来证明,最大的库尔贝克-莱布勒散度的RBM模型与$n$可见和$m$隐藏的单位是有界的$n - \left\lfloor \log(m+1) \right\rfloor - \frac{m+1}{2^{\left\lfloor\log(m+1)\right\rfloor}} \approx (n -1) - \log(m+1)$。通过这种方式,我们可以指定隐藏单元的数量,以确保包含不同类别分布的足够丰富的模型,并遵守给定的容错性。

英文摘要:

We present explicit classes of probability distributions that can be learned by Restricted Boltzmann Machines (RBMs) depending on the number of units that they contain, and which are representative for the expressive power of the model. We use this to show that the maximal Kullback-Leibler divergence to the RBM model with $n$ visible and $m$ hidden units is bounded from above by $n - \left\lfloor \log(m+1) \right\rfloor - \frac{m+1}{2^{\left\lfloor\log(m+1)\right\rfloor}} \approx (n -1) - \log(m+1)$. In this way we can specify the number of hidden units that guarantees a sufficiently rich model containing different classes of distributions and respecting a given error tolerance.

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。

↑