Exact recovery in block spin Ising models at the critical line

被引:3
作者
Loewe, Matthias [1 ]
Schubert, Kristina [2 ]
机构
[1] Univ Munster, Fachbereich Math & Informat, Orleans Ring 10, D-48149 Munster, Germany
[2] Tech Univ Dortmund, Fak Math, Vogelpothsweg 87, D-44227 Dortmund, Germany
来源
ELECTRONIC JOURNAL OF STATISTICS | 2020年 / 14卷 / 01期
关键词
Block models; Ising model; Curie-Weiss model; fluctuations; critical temperature; CURIE-WEISS MODEL; RECONSTRUCTION;
D O I
10.1214/20-EJS1703
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We show how to exactly reconstruct the block structure at the critical line in the so-called Ising block model. This model was recently re-introduced by Berthet, Rigollet and Srivastava in [2]. There the authors show how to exactly reconstruct blocks away from the critical line and they give an upper and a lower bound on the number of observations one needs; thereby they establish a minimax optimal rate (up to constants). Our technique relies on a combination of their methods with fluctuation results obtained in [20]. The latter are extended to the full critical regime. We find that the number of necessary observations depends on whether the interaction parameter between two blocks is positive or negative: In the first case, there are about N log N observations required to exactly recover the block structure, while in the latter case root N log N observations suffice.
引用
收藏
页码:1796 / 1815
页数:20
相关论文
共 23 条
  • [1] Amini A. A., 2018, ANN STAT, V46, P149
  • [2] [Anonymous], 2006, CLASSICS MATH
  • [3] EXACT RECOVERY IN THE ISING BLOCKMODEL
    Berthet, Quentin
    Rigollet, Philippe
    Srivastava, Piyush
    [J]. ANNALS OF STATISTICS, 2019, 47 (04) : 1805 - 1834
  • [4] RIGOROUS BOUNDS ON THE STORAGE CAPACITY OF THE DILUTE HOPFIELD MODEL
    BOVIER, A
    GAYRARD, V
    [J]. JOURNAL OF STATISTICAL PHYSICS, 1992, 69 (3-4) : 597 - 627
  • [5] Efficiently Learning Ising Models on Arbitrary Graphs
    Bresler, Guy
    [J]. STOC'15: PROCEEDINGS OF THE 2015 ACM SYMPOSIUM ON THEORY OF COMPUTING, 2015, : 771 - 782
  • [6] RECONSTRUCTION OF MARKOV RANDOM FIELDS FROM SAMPLES: SOME OBSERVATIONS AND ALGORITHMS
    Bresler, Guy
    Mossel, Elchanan
    Sly, Allan
    [J]. SIAM JOURNAL ON COMPUTING, 2013, 42 (02) : 563 - 578
  • [7] Macroscopic Limit of a Bipartite Curie-Weiss Model: A Dynamical Approach
    Collet, Francesca
    [J]. JOURNAL OF STATISTICAL PHYSICS, 2014, 157 (06) : 1301 - 1319
  • [8] Social distance, heterogeneity and social interactions
    Cont, Rama
    Loewe, Matthias
    [J]. JOURNAL OF MATHEMATICAL ECONOMICS, 2010, 46 (04) : 572 - 590
  • [9] Scaling Limits for Multi-species Statistical Mechanics Mean-Field Models
    Fedele, Micaela
    Contucci, Pierluigi
    [J]. JOURNAL OF STATISTICAL PHYSICS, 2011, 144 (06) : 1186 - 1205
  • [10] Gallo I., 2008, MATH PHYS ELECT J, V14, P21