-
Notifications
You must be signed in to change notification settings - Fork 1
Home
dkolobok edited this page Mar 16, 2017
·
1 revision
Welcome to the confest wiki!
The package confest is developed to execute the identification process of ODE models.
To identify a model is to specialize the parameter values by taking into account experimental data.
The package capabilities include:
- Model fitting. Finds the maximum likelihood estimate for the parameter vector.
- Confidence estimation. Finds confidence intervals for model parameters and confidence bands for model output. Confidence sets can be constructed via the likelihood ratio test method and the bootstrapping method.
- Profile likelihood and identifiability analysis. Constructs likelihood profiles for parameters to test the identifiability of the model.
- Optimal design. Finds the design of experiment that is optimal in the sense of maximizing (minimizing) the design criteria, such as a size of a confidence set.