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 2305.01358cs.DS
收藏

无次子序列性质的基于样本的距离近似

Sample-based distance-approximation for subsequence-freeness

Omer Cohen Sidon, Dana Ron

AI总结:

该研究在基于样本的无分布模型下,提出样本复杂度为$\tilde{O}(k^2/\delta^2)$的算法,近似序列到无次子序列性质的距离,证明了对$1/\delta$的二次依赖是必要的。

摘要:

在本研究中,我们探讨了基于样本的无分布模型下,到无次子序列性质的距离近似问题。对于给定的子序列(词)$w = w_1 \dots w_k$,若存在下标$1 \leq i_1 < \dots < i_k \leq n$使得对所有$1 \leq j \leq k$都有$t_{i_{j}} = w_j$,则称序列(文本)$T = t_1 \dots t_n$包含$w$;否则$T$是$w$-自由的。Ron和Rosin(ACM TOCT 2022)已证明,在基于样本的无分布模型中,对无次子序列性质进行单边错误测试,所需样本数的上下界均为$\Theta(k/\varepsilon)$。我们用$\Delta(T,w,p)$表示在分布$p :[n]\to [0,1]$下,$T$到$w$-自由的距离,目标是得到一个估计值$\widehat\Delta$,使得对于给定的距离参数$\delta$,$|\widehat\Delta - \Delta(T,w,p)| \leq \delta$成立的概率至少为$2/3$。我们的核心结果是一个样本复杂度为$\tilde{O}(k^2/\delta^2)$的算法。我们首先提出一种适用于底层分布$p$为均匀分布的算法,随后说明如何修改该算法使其适用于任意(未知)分布$p$。我们还证明了对$1/\delta$的二次依赖是必要的。

英文摘要:

In this work, we study the problem of approximating the distance to subsequence-freeness in the sample-based distribution-free model. For a given subsequence (word) $w = w_1 \dots w_k$, a sequence (text) $T = t_1 \dots t_n$ is said to contain $w$ if there exist indices $1 \leq i_1 < \dots < i_k \leq n$ such that $t_{i_{j}} = w_j$ for every $1 \leq j \leq k$. Otherwise, $T$ is $w$-free. Ron and Rosin (ACM TOCT 2022) showed that the number of samples both necessary and sufficient for one-sided error testing of subsequence-freeness in the sample-based distribution-free model is $Θ(k/ε)$. Denoting by $Δ(T,w,p)$ the distance of $T$ to $w$-freeness under a distribution $p :[n]\to [0,1]$, we are interested in obtaining an estimate $\widehatΔ$, such that $|\widehatΔ - Δ(T,w,p)| \leq δ$ with probability at least $2/3$, for a given distance parameter $δ$. Our main result is an algorithm whose sample complexity is $\tilde{O}(k^2/δ^2)$. We first present an algorithm that works when the underlying distribution $p$ is uniform, and then show how it can be modified to work for any (unknown) distribution $p$. We also show that a quadratic dependence on $1/δ$ is necessary.

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。

↑