I have a heuristic model of a physical process which is dependent on outcomes from random variables. The physical process approximated by the random number generator suggests what the distribution should be for that generator. For instance, a hypergeometric distribution to govern the introduction of a dispersed phase into a spray. The model takes inputs such as this and then processes those random draws and computes an outcome distribution of droplet sizes. I have experimental measurements of droplet size, and I want to adjust the parameters in the various distributions so that the model more closely predicts the outcome measured experimentally. What is this called? Is it point estimation? Should I have any expectation of robustness under, say changing surface tension? What should I be readying on this subject?
Thanks in advance!