Truncated Gaussian Distribution
A probability distribution obtained by restricting a multivariate Gaussian distribution to a convex polyhedral feasible region. Its moments lack closed-form analytical expressions and require numerical methods to compute. Commonly used in Bayesian inference problems subject to inequality constraints.
Timeline (last 90 days)
The distribution obtained from joint conditioning is no longer Gaussian, but remains computable via truncated-Gaussian moments.
The moments of the truncated Gaussian distribution have no closed-form analytical expression and are typically computed using methods such as quasi-Monte Carlo sampling, the Mini-max tilting algorithm, or variational approximations.
All Facts (2)
The distribution obtained from joint conditioning is no longer Gaussian, but remains computable via truncated-Gaussian moments.
50%UnverifiedThe moments of the truncated Gaussian distribution have no closed-form analytical expression and are typically computed using methods such as quasi-Monte Carlo sampling, the Mini-max tilting algorithm, or variational approximations.
50%