Abstract

The measles the Semarang experience fluctuates every year, so that the City Health Agency (DKK) Semarang put special attention to reducing many cases measles.In the case of smallpox semarang 2013 was data discrete Poisson. Regression Poisson is nonlinear regression used to analyze data count variable response Poisson and meet the equidispersi. In practice often occurs in violation of discrete overdispersi analysis of data in regression poisson underdispersi and models or improper use.To anticipate such violation used Generalized Poisson Regression in modeling (GPR) data. In this research are variable response used in the case of smallpox Semarang 2013 and variable prediktor used is many medicines measles, community health centers, the poverty and overcrowding every subdistrict across Semarang town. The best model Generalized Poisson Regression (GPR) was gotten.