English

Recursive computation for evaluating the exact $p$-values of temporal and spatial scan statistics

Computation 2015-11-03 v1 Methodology

Abstract

Let VV be a finite set of indices, and let BiB_i, i=1,,mi=1,\ldots,m, be subsets of VV such that V=i=1mBiV=\bigcup_{i=1}^{m}B_i. Let XiX_i, iVi\in V, be independent random variables, and let XBi=(Xj)jBiX_{B_i}=(X_j)_{j\in B_i}. In this paper, we propose a recursive computation method to calculate the conditional expectation E[i=1mχi(XBi)N]E\bigl[\prod_{i=1}^m\chi_i(X_{B_i}) \,|\, N\bigr] with N=iVXiN=\sum_{i\in V}X_i given, where χi\chi_i is an arbitrary function. Our method is based on the recursive summation/integration technique using the Markov property in statistics. To extract the Markov property, we define an undirected graph whose cliques are BjB_j, and obtain its chordal extension, from which we present the expressions of the recursive formula. This methodology works for a class of distributions including the Poisson distribution (that is, the conditional distribution is the multinomial). This problem is motivated from the evaluation of the multiplicity-adjusted pp-value of scan statistics in spatial epidemiology. As an illustration of the approach, we present the real data analyses to detect temporal and spatial clustering.

Cite

@article{arxiv.1511.00108,
  title  = {Recursive computation for evaluating the exact $p$-values of temporal and spatial scan statistics},
  author = {Satoshi Kuriki and Kunihiko Takahashi and Hisayuki Hara},
  journal= {arXiv preprint arXiv:1511.00108},
  year   = {2015}
}

Comments

23 pages, 7 figures, 3 tables

R2 v1 2026-06-22T11:33:43.460Z