Maximum a Posteriori Estimation by Search in Probabilistic Programs
This addresses the need for efficient MAP estimation in probabilistic programs, which is incremental as it builds on existing algorithms with improvements in speed and robustness.
The paper tackles the problem of fast maximum a posteriori probability estimation in probabilistic programs by introducing Bayesian ascent Monte Carlo (BaMC), an approximate search algorithm that is faster and more robust than other methods on a range of models.
We introduce an approximate search algorithm for fast maximum a posteriori probability estimation in probabilistic programs, which we call Bayesian ascent Monte Carlo (BaMC). Probabilistic programs represent probabilistic models with varying number of mutually dependent finite, countable, and continuous random variables. BaMC is an anytime MAP search algorithm applicable to any combination of random variables and dependencies. We compare BaMC to other MAP estimation algorithms and show that BaMC is faster and more robust on a range of probabilistic models.