Conditional Approximate Normalizing Flows for Joint Multi-Step Probabilistic Forecasting with Application to Electricity Demand

TL;DR

This paper introduces conditional approximate normalizing flows (CANF), a novel method for probabilistic multi-step time-series forecasting that captures long-term correlations, demonstrated on electricity demand data with significant improvements in accuracy and decision-making.

Contribution

The paper presents CANF, a new approach that effectively models long-term dependencies in probabilistic forecasts, outperforming existing methods in accuracy and scheduling applications.

Findings

CANF reduces KL divergence by one-third on toy data.

CANF outperforms other methods in multi-step forecast accuracy.

CANF enables up to 10x better scheduling decisions.

Abstract

Some real-world decision-making problems require making probabilistic forecasts over multiple steps at once. However, methods for probabilistic forecasting may fail to capture correlations in the underlying time-series that exist over long time horizons as errors accumulate. One such application is with resource scheduling under uncertainty in a grid environment, which requires forecasting electricity demand that is inherently noisy, but often cyclic. In this paper, we introduce the conditional approximate normalizing flow (CANF) to make probabilistic multi-step time-series forecasts when correlations are present over long time horizons. We first demonstrate our method's efficacy on estimating the density of a toy distribution, finding that CANF improves the KL divergence by one-third compared to that of a Gaussian mixture model while still being amenable to explicit conditioning. We then use a publicly available household electricity consumption dataset to showcase the effectiveness of CANF on joint probabilistic multi-step forecasting. Empirical results show that conditional approximate normalizing flows outperform other methods in terms of multi-step forecast accuracy and lead to up to 10x better scheduling decisions. Our implementation is available at https://github.com/sisl/JointDemandForecasting.