3 ms·

Thanks. This helps. Can you give a specific example of when someone would use this calculation, and where it would go wrong? Or what kind of problem requires

Thanks. This helps. Can you give a specific example of when someone would use this calculation, and where it would go wrong? Or what kind of problem requires the draws to be a function of all the data?

tristanz 13y ago
This is the core calculation to all of Bayesian statistics, so it's used pretty much everywhere. Almost any time you try to compute p(unknown|data) you run into computational issues. It's a high dimensional integration problem that has no closed form solution. There are models that can be simplified, but they are the minority.