search

gaurav-arya / StochasticAD.jl

Research package for automatic differentiation of programs containing discrete randomness.
MIT License
198 stars 16 forks source link

Fixes to Poisson autoregression #37

Closed mschauer closed 1 year ago

mschauer commented 1 year ago

Replaces #31

mschauer commented 1 year ago

Maximising the marginal posterior density makes sense.

I think we could also maximise the expected log-likelihood.

But that would require a bit of work to check for validity.

mschauer commented 1 year ago

I convinced my self that short of doing expectation maximisation we cannot use the log likelihood just like this.