A probability density is log-concave when its logarithm is a concave function on its support. Tangent lines to a differentiable concave log density lie above its graph, giving exponential envelopes for rejection sampling.
Articles by others on the same topic
There are currently no matching articles.