Multivariate Zero-Inflated Poisson Models and Their Applications

Li, Chin‐Shang; Lu, Jye‐Chyi; Park, Jinho; Kim, Kyungmoo; Brinkley, Paul A.; Peterson, J.P.

doi:10.2307/1270992

Cited by 77 publications

(97 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Lambert (1992) provided techniques for modeling data with a zero-inflated Poisson (ZIP) model. Incorporating zero-inflated likelihoods in spatially explicit models is described in Agarwal et al (2002), Barry and Welsh (2002), Fahrmeir and Echavarria (in press), Gschlobl and Czado (2006), Rathbun and Fei (2006), Rigby and Stasinopoulos (2005) and in general (not spatial) in Hall (2000), Lambert (1992), and Li et al (1999 SemiPar via a space-filling algorithm (Ruppert et al 2003) with the default number of 50 knots.…”

Section: Methodsmentioning

confidence: 99%

Proceedings of the eighth annual forest inventory and analysis symposium

McRoberts¹,

Reams²,

Deusen³

et al. 2009

View full text Add to dashboard Cite

Section: Methodsmentioning

confidence: 99%

Proceedings of the eighth annual forest inventory and analysis symposium

McRoberts¹,

Reams²,

Deusen³

et al. 2009

View full text Add to dashboard Cite

“…The bivariate zero-inflated Poisson (BZIP) regression is an extension of the univariate ZIP for the joint analysis of positively correlated counts with excess zeros. Li et al (1999), Wang et al (2003), Karlis and Ntzoufras (2005) and Lee et al (2005) give formal descriptions of the model and its estimation process. In what follows, we shall discuss the generic structure of a BZIP model in the light of our substantive research questions.…”

Section: Statistical Modelmentioning

confidence: 99%

Concentration of personal and household crimes in England and Wales

et al. 2010

View full text Add to dashboard Cite

Crime is disproportionally concentrated in few areas. Though long-established, there remains uncertainty about the reasons for variation in the concentration of similar crime (repeats) or different crime (multiples). Wholly neglected have been composite crimes when more than one crime types coincide as parts of a single event. The research reported here disentangles area crime concentration into repeats, multiple and composite crimes. The results are based on estimated bivariate zero-inflated Poisson regression models with covariance structure which explicitly account for crime rarity and crime concentration. The implications of the results for criminological theorizing and as a possible basis for more equitable police funding are discussed.

show abstract

“…Since Li et al (1999) proposed the zero-inflated bivariate count regression models, Walhin (2001) and Wang et al (2003) showed the several zero inflated Poisson, denoted by BZIP models, could be extended to the zero inflated bivariate models using various Poisson mixtures. But, it is often for the zero-inflation to connect with the overdispersion, and it is more natural to use negative binomial distributions instead of Poisson distributions.…”

Section: Introductionmentioning

confidence: 99%