Timezone: »

 
Spotlight
Calibrated and Sharp Uncertainties in Deep Learning via Density Estimation
Volodymyr Kuleshov · Shachi Deshpande

Thu Jul 21 01:50 PM -- 01:55 PM (PDT) @ Room 310
in Optimization/Probabilistic Methods »

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.

Keywords: Deep Learning · Probabilistic Methods · PM: Bayesian Models and Methods · MISC: Supervised Learning

Author Information

Volodymyr Kuleshov (Cornell University)
Shachi Deshpande (Cornell University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors