Inverse Problems
An inverse problem is a collection of trials which define a multi-simulation optimization problem. The solution to an inverse problem are the "good parameters" which make the simulations simultaniously fit their respective data sufficiently well.
Constructing InverseProblems
JuliaSimModelOptimizer.InverseProblem — TypeInverseProblem(trials, model, search_space)The InverseProblem stores the information needed to perform the optimization and find the model parameters and/or initial conditions that best fit the data. The model is an MTK ODESystem describing the dynamics of the investigated system. The trials constitute a collection that groups the individual trials according to their interpretation, such as IndependentTrials (there is no link between trials) or SteadyStateTrials (the first trial computes a steady state that other trials may continue to solve from).
The search_space is a Vector of pairs. Each pair consists of a parameter or initial condition to be optimized and its lower and upper bounds. When using an MCMCOpt method, search_space can hold distributions for each parameter instead of bounds, e.g. see TransformedBeta or see the Distributions.jl documentation) for more options.