Temporal Probability Calibration

TL;DR

This paper addresses the challenge of calibrating class probability estimates from sequential data, proposing adaptive methods that improve calibration for incomplete sequences in various applications.

Contribution

It introduces novel calibration techniques tailored for sequential models with incomplete data, outperforming traditional methods in effectiveness.

Findings

Proposed methods significantly improve probability calibration accuracy.

Adaptive calibration schemes outperform traditional techniques on sequential data.

Effective across multiple application domains.

Abstract

In many applications, accurate class probability estimates are required, but many types of models produce poor quality probability estimates despite achieving acceptable classification accuracy. Even though probability calibration has been a hot topic of research in recent times, the majority of this has investigated non-sequential data. In this paper, we consider calibrating models that produce class probability estimates from sequences of data, focusing on the case where predictions are obtained from incomplete sequences. We show that traditional calibration techniques are not sufficiently expressive for this task, and propose methods that adapt calibration schemes depending on the length of an input sequence. Experimental evaluation shows that the proposed methods are often substantially more effective at calibrating probability estimates from modern sequential architectures for incomplete sequences across a range of application domains.