arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 0812.0933cs.LGcs.CC
收藏

决策树在大多数产品分布中是PAC可学习的:平滑分析

Decision trees are PAC-learnable from most product distributions: a smoothed analysis

Microsoft Research New England(新英格兰微软研究院);Boston University(波士顿大学)

Adam Tauman Kalai, Shang-Hua Teng

摘要:

我们考虑PAC学习决策树的问题,即,从独立的随机标记的示例学习n维超立方体上的决策树。尽管付出了巨大的努力,但没有已知的多项式时间算法用于学习多项式大小的决策树(甚至是任何超常数大小的树),即使假设示例是从{0,1}^n上的均匀分布中提取的。我们给出了一个算法,学习任意多项式大小的决策树{ \em大多数产品分布}。特别地,考虑一个随机乘积分布,其中每个比特的偏置是独立且均匀地从[.49,.51]中选择的。然后,该算法将以产品分布参数和从中提取的随机示例的高概率学习任何树。更一般地说,在平滑分析的精神,我们考虑一个任意的产品分布,其参数被指定只有一个[-c,c]精度(扰动),为任意小的正常数c。

英文摘要:

We consider the problem of PAC-learning decision trees, i.e., learning a decision tree over the n-dimensional hypercube from independent random labeled examples. Despite significant effort, no polynomial-time algorithm is known for learning polynomial-sized decision trees (even trees of any super-constant size), even when examples are assumed to be drawn from the uniform distribution on {0,1}^n. We give an algorithm that learns arbitrary polynomial-sized decision trees for {\em most product distributions}. In particular, consider a random product distribution where the bias of each bit is chosen independently and uniformly from, say, [.49,.51]. Then with high probability over the parameters of the product distribution and the random examples drawn from it, the algorithm will learn any tree. More generally, in the spirit of smoothed analysis, we consider an arbitrary product distribution whose parameters are specified only up to a [-c,c] accuracy (perturbation), for an arbitrarily small positive constant c.

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。

↑