To detect the effect of count data with many zeros on the zeroinflated poisson regression model, the authors studied the global diagnostics analysis and using the completedata log-likelihood function and the EM algorithm. The one-step approximation of parameter estimates and some statistics, such as generalized cook distance, Q-distance, are obtained. Meantime, the generalized leverage was discussed based on the EM algorithm. Finally, a numerical example was given to illustrate our results.
LI Ai-ping, GU Zheng, XIE Feng-chang.
Case Deletion Measures and Generalized Leverage for the Zero Inflated Poisson Regression Models[J]. JOURNAL OF NANJING FORESTRY UNIVERSITY. 2007, 31(06): 109-112 https://doi.org/10.3969/j.jssn.1000-2006.2007.06.026