The simultaneous two-step entropy model
The simultaneous two-stage entropy model is: \[ \min_{p,z}G(p,q(z))+alpha * G(q(z),q^ 0) \] where \(p\geq 0\) and other linear constraints in p; q(z) is a function from \(R^ k\) to \(\{\) x: \(x\geq 0\), \(x\in R^ n\}\); \(q^ 0\geq 0\); alpha is a non-negative constant. The proposed simultaneous two-step entropy model seems to be a very useful approach to use data from more than one observation to estimate a forecast. A user is able to formulate his own q(z), which is significant for his specific application. The model can also be used for analyzing; then, the vector Z should be considered as a vector of parameters and it is this vector which is of main interest when the two-step entropy problem is to be solved. A programme package, commercially applied for migration and traffic planning forecast problems is also pointed out.
This page was built for publication: The simultaneous two-step entropy model
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1120474)
