coconots

CRANcoconotsMaRDI QIDQ60267

Convolution-Closed Models for Count Time Series

Robert C. Jung, Andy R. Tremayne, Manuel Huth

Last update: 1 October 2023

Copyright license: MIT license, File License

Software version identifier: 1.1.1, 1.1.2, 1.1, 1.1.3

Description

Useful tools for fitting, validating, and forecasting of practical convolution-closed time series models for low counts are provided. Marginal distributions of the data can be modeled via Poisson and Generalized Poisson innovations. Regression effects can be modelled via time varying innovation rates. The models are described in Jung and Tremayne (2011) <doi:10.1111/j.1467-9892.2010.00697.x> and the model assessment tools are presented in Czado et al. (2009) <doi:10.1111/j.1541-0420.2009.01191.x>, Gneiting and Raftery (2007) <doi:10.1198/016214506000001437> and, Tsay (1992) <doi:10.2307/2347612>.

Cites work

• Convolution-closed models for count time series with applications
• Predictive Model Assessment for Count Data
• Strictly Proper Scoring Rules, Prediction, and Estimation
• Model Checking via Parametric Bootstraps in Time Series Analysis

This page was built for software: coconots