Wednesday, June 1, 2011

Parameters, models and data

Let's take some interesting data.
Data plot does not look too complicated.
A two variable, two parameter model can explain it mathematically.
Let's make a model.
With around 50 parameters and a dozen equations.
Let's fit the data to the model, with some algorithm.
For some parameter set, it will surely fit.

Let's publish.

If someone asks something, tell: the model is right as long as it can explain data.

Repeatability? Sensitivity?

Next paper, may be...

But sorry, we dont have funding, so no next paper on the topic.


No comments: