Related Experiment Videos
NHPPREG: a computer program for the analysis of nonhomogeneous Poisson process data with covariates
1McGill University, Montreal Children's Hospital Research Institute, Canada.
Computer Methods and Programs in Biomedicine
|June 1, 1992
Summary
This study introduces a C program for estimating parameters in Poisson point processes with covariates. It enables both parametric and semi-parametric modeling, incorporating random effects for extra Poisson variation.
Area of Science:
- Statistical Modeling
- Computational Statistics
Background:
- Point processes are fundamental in analyzing event data.
- Modeling these processes often requires estimating complex parameters.
- Existing methods may lack flexibility in handling covariate effects and extra variation.
Purpose of the Study:
- To develop a computational tool for parameter estimation in point processes.
- To implement maximum likelihood estimation for regression coefficients.
- To accommodate both parametric and semi-parametric modeling strategies.
Main Methods:
- A C program was developed for statistical analysis.
- Maximum likelihood estimation (MLE) was employed.
- The program supports parametric, semi-parametric, and random effects models for Poisson point processes.
Main Results:
- The program successfully estimates regression coefficients and parameters for point processes.
- It provides flexibility in model choice (parametric vs. semi-parametric).
- The inclusion of random effects allows for modeling extra Poisson variation.
Conclusions:
- The developed C program offers a robust and flexible tool for analyzing point process data.
- It facilitates advanced statistical modeling, including the incorporation of covariates and random effects.
- This work provides a practical implementation based on Lawless (1987) theory.