Model Calibration

I’m spending all my time calibrating my model, lately. “What is that?” you ask. It’s a process in which I modify my model such that the model output data is statistically similar to observed data. In this case, I’m trying to make all the dots on this chart land right where the big blue dots are. I’ve mapped 3 replications so far, each of which is statistically and significantly different from the observed data i’m trying to match. ….but it’s probably a little unrealistic to expect all of my replications to be similar to the observed data. What really matters is that the average of all three (or more as time allows) replications is similar to the observed data. Hopefully, t-score less than 2, but probably, I’ll call it good if it’s within one Standard Deviation. I’m not very good at calibrating. I’m blaming it on malnutrition. I wish i had some lasagna right now.

Comments are closed.