11 Variational Inference Algorithm: ADVI
CmdStan can approximate the posterior distribution using variational inference. The approximation is a Gaussian in the unconstrained variable space.
Stan implements an automatic variational inference algorithm, called Automatic Differentiation Variational Inference (ADVI) Kucukelbir et al. (2015). ADVI uses Monte Carlo integration to approximate the variational objective function, the ELBO (evidence lower bound). ADVI optimizes the ELBO in the real-coordinate space using stochastic gradient ascent. The measures of convergence are similar to the relative tolerance scheme of Stan’s optimization algorithms.
The algorithm progression consists of an adaptation phase followed by a sampling phase.
The adaptation phase finds a good value for the step size scaling parameter eta
.
The evidence lower bound (ELBO) is the variational objective function and is evaluated based on a Monte Carlo estimate.
The variational inference algorithm in Stan is stochastic,
which makes it challenging to assess convergence.
The algorithm runs until the mean change in ELBO drops below
the specified tolerance.
The full set of configuration options available for the variational
method is
reported at the beginning of the sampler output file as CSV comments.
When the example model bernoulli.stan
is run with method=variational
via the command line with all default arguments,
the resulting Stan CSV file header comments show the complete set
of default configuration options:
# method = variational
# variational
# algorithm = meanfield (Default)
# meanfield
# iter = 10000 (Default)
# grad_samples = 1 (Default)
# elbo_samples = 100 (Default)
# eta = 1 (Default)
# adapt
# engaged = 1 (Default)
# iter = 50 (Default)
# tol_rel_obj = 0.01 (Default)
# eval_elbo = 100 (Default)
# output_samples = 1000 (Default)
The console output includes a notice that this algorithm is considered to be experimental:
EXPERIMENTAL ALGORITHM:
This procedure has not been thoroughly tested and may be unstable
or buggy. The interface is subject to change.
11.1 Variational algorithms
Stan implements two variational algorithms.
The algorithm
argument specifies the variational algorithm.
algorithm=meanfield
- Use a fully factorized Gaussian for the approximation. This is the default algorithm.algorithm=fullrank
Use a Gaussian with a full-rank covariance matrix for the approximation.
11.2 Configuration
iter=<int>
Maximum number of iterations. Must be \(> 0\). Default is \(10000\).grad_samples=<int>
Number of samples for Monte Carlo estimate of gradients. Must be \(> 0\). Default is \(1\).elbo_samples=<int>
Number of samples for Monte Carlo estimate of ELBO (objective function). Must be \(> 0\). Default is \(100\).eta=<double>
Stepsize weighting parameter for adaptive stepsize sequence. Must be \(> 0\). Default is \(1.0\).adapt
Warmup Adaptation keyword, takes sub-arguments:engaged=<boolean>
Adaptation engaged? Valid values: \((0, 1)\). Default is \(1\).iter=<int>
Maximum number of adaptation iterations. Must be \(> 0\). Default is \(50\).
tol_rel_obj=<double>
Convergence tolerance on the relative norm of the objective. Must be \(> 0\). Default is \(0.01\).eval_elbo=<int>
Evaluate ELBO every Nth iteration. Must be \(> 0\). Default is 100.output_samples=<int>
Number of posterior samples to draw and save. Must be \(> 0\). Default is 1000.
11.3 CSV output
The output file consists of the following pieces of information:
The full set of configuration options available for the
variational
method is reported at the beginning of the sampler output file as CSV comments.The first three output columns are labelled
lp__
,log_p__
,log_g__
, the rest are the model parameters.The stepsize adaptation information is output as CSV comments following column header row.
The following line contains the mean of the variational approximation.
The rest of the output contains
output_samples
number of samples drawn from the variational approximation.
To illustrate, we call Stan’s variational inference on the example model and data:
> ./bernoulli variational data file=bernoulli.data.R
By default, the output file is output.csv
.
Lines 1 - 28 contain configuration information:
# stan_version_major = 2
# stan_version_minor = 23
# stan_version_patch = 0
# model = bernoulli_model
# method = variational
# variational
# algorithm = meanfield (Default)
# meanfield
# iter = 10000 (Default)
# grad_samples = 1 (Default)
# elbo_samples = 100 (Default)
# eta = 1 (Default)
# adapt
# engaged = 1 (Default)
# iter = 50 (Default)
# tol_rel_obj = 0.01 (Default)
# eval_elbo = 100 (Default)
# output_samples = 1000 (Default)
...
The column header row is:
lp__,log_p__,log_g__,theta
The stepsize adaptation information is:
# Stepsize adaptation complete.
# eta = 1
The reported mean variational approximations information is:
0,0,0,0.214911
That is, the estimate for theta
given the data is 0.2
.
The following is a sample based on this approximation:
0,-14.0252,-5.21718,0.770397
0,-7.05063,-0.10025,0.162061
0,-6.75031,-0.0191099,0.241606
...