Re: [math-fun] Curve-fitting methods ?

Yes, this is the kind of posterior model checking that we often forget to do. - Cris

On Sep 23, 2018, at 6:06 PM, Henry Baker <hbaker1@pipeline.com> wrote:

One thing this paper does seem to recommend: estimate your model & parameters, then generate synthetic data from that model with those parameters. If the resulting dataset doesn't look at all like your original dataset, you're barking up the wrong tree!