首页> 外文会议>STACS 98 >Random Sparse Bit Strings at the Threshold of Adjacency
【24h】

Random Sparse Bit Strings at the Threshold of Adjacency

机译:相邻阈值处的随机稀疏位字符串

获取原文
获取原文并翻译 | 示例

摘要

We give a complete characterization for the limit probabilities of first order sentences over sparse random bit strings at the threshold of adjacency. For strings of length n, we let the probability that a bit is "on" be c/square root of (n), fro a real positive number c. For the every first order sentence fai, we show that the limit probability function: f sub fai(c)=lim sub n-infinity Pr[U sub n, c/square root of (n) has the property fai] is infinitely differnetiable. Our methodology for showing this is in itself interesting. We begin with finite models, go to the infinite (via the almost sure theories) and then characterize f sub fai(c) as an infinite sum of products ofpolynomials and exponentials. We further show that if a sentence fai has limiting probability 1 for some c, then fai has limiting probability idnetically 1 for every c. This gives the surprising result that the almost sure theories are idnetical for every c.
机译:我们给出了稀疏随机位串上邻接阈值处一阶句子的极限概率的完整表征。对于长度为n的字符串,我们让一个位“在”上的概率为(n)的c /平方根,即实数正数c。对于每一个一阶句子fai,我们证明极限概率函数:f sub fai(c)= lim sub n-infinity Pr [U sub n,(n)的c /平方根具有fai]的性质是无限可微的。我们展示这一点的方法本身很有趣。我们从有限模型开始,进入无限(通过几乎确定的理论),然后将f sub fai(c)表征为多项式和指数乘积的无限和。我们进一步证明,如果一个句子fai对于某个c具有极限概率1,那么fai在每个c上具有固有的极限概率1。这给出了令人惊讶的结果,即几乎可以肯定的理论对于每个c都是理论上的。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号