Regularized estimating equations for model selection of clustered spatial point processes

Andrew L. Thurman, Rao Fu, Yongtao Guan, Jun Zhu

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Clustered spatial point processes are popular models for spatial point pattern data that contain clusters of events. For regression purposes, however, statistically rigorous methods for model selection appear to be lacking and are the focus of this paper. Here, an unbiased estimating equation is considered for parameter estimation to simplify computation and in addition, a weighted estimating equation is adopted to improve statistical efficiency. In particular, both regularized unweighted and regularized weighted estimating equations are developed for simultaneous variable selection and parameter estimation. Asymptotic properties of the proposed method are established and finite sample properties are assessed in a simulation study. For illustration, our method is applied to evaluate and quantify the relationship between the locations of a tropical tree species and over 200 covariates in a forest plot on the Barro Colorado Island.

Original languageEnglish (US)
Pages (from-to)173-188
Number of pages16
JournalStatistica Sinica
Issue number1
StatePublished - Jan 2015


  • Ecological application
  • Spatial interaction
  • Spatial point patterns
  • Spatial statistics
  • Variable selection
  • Weighted estimating equations

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty


Dive into the research topics of 'Regularized estimating equations for model selection of clustered spatial point processes'. Together they form a unique fingerprint.

Cite this