Uncertainty Prediction for Deep Sequential Regression Using Meta Models

TL;DR

This paper introduces a flexible meta-modeling approach for deep sequential regression that provides accurate, non-stationarity-agnostic uncertainty estimates, outperforming existing methods in real-world scenarios.

Contribution

It presents a novel meta-modeling technique that generates symmetric and asymmetric uncertainty estimates without assuming stationarity, improving practical sequential regression applications.

Findings

Outperforms baseline methods in drift and non-drift scenarios

Provides symmetric and asymmetric uncertainty estimates

No assumptions about stationarity required

Abstract

Generating high quality uncertainty estimates for sequential regression, particularly deep recurrent networks, remains a challenging and open problem. Existing approaches often make restrictive assumptions (such as stationarity) yet still perform poorly in practice, particularly in presence of real world non-stationary signals and drift. This paper describes a flexible method that can generate symmetric and asymmetric uncertainty estimates, makes no assumptions about stationarity, and outperforms competitive baselines on both drift and non drift scenarios. This work helps make sequential regression more effective and practical for use in real-world applications, and is a powerful new addition to the modeling toolbox for sequential uncertainty quantification in general.