Shift Happens: Adjusting Classifiers
원문 초록
Minimizing expected loss measured by a proper scoring rule, such as Brier score or log-loss (cross-entropy), is a common objective while training a probabilistic classifier. If the data have experienced dataset shift where the class distributions change post-training, then often the model's performance will decrease, over-estimating the probabilities of some classes while under-estimating the others on average. We propose unbounded and bounded general adjustment (UGA and BGA) methods that transform all predictions to (re-)equalize the average prediction and the class distribution. These methods act differently depending on which proper scoring rule is to be minimized, and we have a theoretical guarantee of reducing loss on test data, if the exact class distribution is known. We also demonstrate experimentally that, when in practice the class distribution is known only approximately, there is often still a reduction in loss depending on the amount of shift and the precision to which the class distribution is known.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
h-calibration: Rethinking Classifier Recalibration with Probabilistic Error-Bounded Objective
Enforcing tail calibration when training probabilistic forecast models
Harnessing Superclasses for Learning from Hierarchical Databases
Bayesian Autoregressive Online Change-Point Detection with Time-Varying Parameters
Improving probabilistic forecasts of extreme wind speeds by training statistical post-processing models with weighted scoring rules
Distributional Regression U-Nets for the Postprocessing of Precipitation Ensemble Forecasts
토론
이 논문과 연결된 포럼 글이 아직 없습니다.