ComputeBetasAndNuHat: Estimate COM-Poisson parameters (betas and nu)
Description
Estimates maximum likelihood estimates for betas and nu using nlminb.
Usage
ComputeBetasAndNuHat(x, y, betainit, nuinit, max)
Arguments
x
matrix of size nxp
y
column vector of length n
betainit
initial vector of betas, b0_1, ..., b0_p
nuinit
initial value for nu
max
maximum number to use for truncating infinite sums
Value
Maximum likelihood estimates for betas and nu using constrained optimization (nlminb).
Details
ComputeBetasAndNuHat finds the maximum likelihood estimates for the COM-Poisson estimators (betas and nu) using constrained optimization (nlminb). The dispersion parameter, nu, is greater than or equal to zero. n=number of records and p=number of predictors.
References
A Flexible Regression Model for Count Data, by Sellers & Shmueli, http://ssrn.com/abstract=1127359