Calibrated and Sharp Uncertainties in Deep Learning via Density Estimation
원문 초록
Accurate probabilistic predictions can be characterized by two properties -- calibration and sharpness. However, standard maximum likelihood training yields models that are poorly calibrated and thus inaccurate -- a 90% confidence interval typically does not contain the true outcome 90% of the time. This paper argues that calibration is important in practice and is easy to maintain by performing low-dimensional density estimation. We introduce a simple training procedure based on recalibration that yields calibrated models without sacrificing overall performance; unlike previous approaches, ours ensures the most general property of distribution calibration and applies to any model, including neural networks. We formally prove the correctness of our procedure assuming that we can estimate densities in low dimensions and we establish uniform convergence bounds. Our results yield empirical performance improvements on linear and deep Bayesian models and suggest that calibration should be increasingly leveraged across machine learning.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Binned semiparametric Bayesian networks
EBC-ZIP: Improving Blockwise Crowd Counting with Zero-Inflated Poisson Regression
Scaling-Up the Pretraining of the Earth Observation Foundation Model PhilEO to the MajorTOM Dataset
FunDiff: Diffusion Models over Function Spaces for Physics-Informed Generative Modeling
TabPFN: One Model to Rule Them All?
Querying Kernel Methods Suffices for Reconstructing their Training Data
토론
이 논문과 연결된 포럼 글이 아직 없습니다.