The Annals of Applied Probability

Thresholds for detecting an anomalous path from noisy environments

Shirshendu Chatterjee and Ofer Zeitouni

Full-text: Access denied (no subscription detected)

We're sorry, but we are unable to provide you with the full text of this article because we are not able to identify you as a subscriber. If you have a personal subscription to this journal, then please login. If you are already logged in, then you may need to update your profile to register your subscription. Read more about accessing full-text


We consider the “searching for a trail in a maze” composite hypothesis testing problem, in which one attempts to detect an anomalous directed path in a lattice 2D box of side $n$ based on observations on the nodes of the box. Under the signal hypothesis, one observes independent Gaussian variables of unit variance at all nodes, with zero mean off the anomalous path and mean $\mu_{n}$ on it. Under the null hypothesis, one observes i.i.d. standard Gaussians on all nodes. Arias-Castro et al. [Ann. Statist. 36 (2008) 1726–1757] showed that if the unknown directed path under the signal hypothesis has known initial location, then detection is possible (in the minimax sense) if $\mu_{n}\gg1/\sqrt{\log n}$, while it is not possible if $\mu_{n}\ll1/\log n\sqrt{\log\log n}$. In this paper, we show that this result continues to hold even when the initial location of the unknown path is not known. As is the case with Arias-Castro et al. [Ann. Statist. 36 (2008) 1726–1757], the upper bound here also applies when the path is undirected. The improvement is achieved by replacing the linear detection statistic used in Arias-Castro et al. [Ann. Statist. 36 (2008) 1726–1757] with a polynomial statistic, which is obtained by employing a multiscale analysis on a quadratic statistic to bootstrap its performance. Our analysis is motivated by ideas developed in the context of the analysis of random polymers in Lacoin [Comm. Math. Phys. 294 (2010) 471–503].

Article information

Ann. Appl. Probab., Volume 28, Number 5 (2018), 2635-2663.

Received: May 2017
Revised: September 2017
First available in Project Euclid: 28 August 2018

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 62C20: Minimax procedures 62G10: Hypothesis testing
Secondary: 82B20: Lattice systems (Ising, dimer, Potts, etc.) and systems on graphs 60K37: Processes in random environments

Detecting a chain of nodes in a network minimax detection random scenery


Chatterjee, Shirshendu; Zeitouni, Ofer. Thresholds for detecting an anomalous path from noisy environments. Ann. Appl. Probab. 28 (2018), no. 5, 2635--2663. doi:10.1214/17-AAP1356.

Export citation


  • [1] Arias-Castro, E., Candès, E. J., Helgason, H. and Zeitouni, O. (2008). Searching for a trail of evidence in a maze. Ann. Statist. 36 1726–1757.
  • [2] Berger, Q. and Lacoin, H. (2017). The high-temperature behavior for the directed polymer in dimension $1+2$. Ann. Inst. Henri Poincaré Probab. Stat. 53 430–450.
  • [3] Gamkrelidze, N. G. and Rotar’, V. I. (1978). The rate of convergence in a limit theorem for quadratic forms. Theory Probab. Appl. 22 394–397.
  • [4] Götze, F. and Tikhomirov, A. N. (1999). Asymptotic distribution of quadratic forms. Ann. Probab. 27 1072–1098.
  • [5] Lacoin, H. (2010). New bounds for the free energy of directed polymers in dimension $1+1$ and $1+2$. Comm. Math. Phys. 294 471–503.
  • [6] Rotar’, V. I. (1974). Certain limit theorems for polynomials of degree two. Theory Probab. Appl. 18 499–507.
  • [7] Rotar’, V. I. and Shervashidze, T. L. (1986). Some estimates for distributions of quadratic forms. Theory Probab. Appl. 30 585–590.
  • [8] Whittle, P. (1960). Bounds for the moments of linear and quadratic forms in independent variables. Theory Probab. Appl. 5 302–305.