Accelerating Generalized Linear Models by Trading off Computation for Uncertainty

Bayesian Generalized Linear Models (GLMs) define a flexible probabilistic framework to model categorical, ordinal and continuous data, and are widely used in practice. However, exact inference in GLMs is prohibitively expensive for large datasets, thus requiring approximations in practice. The resulting approximation error adversely impacts the reliability of the model and is not accounted for in the uncertainty of the prediction. In this work, we introduce a family of iterative methods that explicitly model this error. They are uniquely suited …

approximation bayesian categorical computation continuous cs.lg data datasets error framework generalized impacts inference large datasets linear ordinal practice reliability stat.ml trading uncertainty