Non-Parametric Calibration of Probabilistic Regression
This work addresses the under-explored issue of calibration in regression, which is important for improving reliability in domains like finance or healthcare, though it appears incremental as it adapts classification methods to regression.
The paper tackled the problem of calibrating probabilistic regression models to improve probability density estimates for real-valued targets, proposing three non-parametric approaches that experimentally enhanced predictive likelihood performance.
The task of calibration is to retrospectively adjust the outputs from a machine learning model to provide better probability estimates on the target variable. While calibration has been investigated thoroughly in classification, it has not yet been well-established for regression tasks. This paper considers the problem of calibrating a probabilistic regression model to improve the estimated probability densities over the real-valued targets. We propose to calibrate a regression model through the cumulative probability density, which can be derived from calibrating a multi-class classifier. We provide three non-parametric approaches to solve the problem, two of which provide empirical estimates and the third providing smooth density estimates. The proposed approaches are experimentally evaluated to show their ability to improve the performance of regression models on the predictive likelihood.