Predicting Electricity Consumption with Random Walks on Gaussian Processes
Keywords: Gaussian Processes, Bayesian decision-making, time-series forecasting, electricity consumption
TL;DR: Domino, a random walk on GPs for time-series forecasting, mitigates the computational cost of large data and over-reliance on specific models in predictive settings, and instantiate it on the problem of short-term electricity forecasting.
Abstract: We consider time-series forecasting problems where data is scarce, difficult to gather, or induces a prohibitive computational cost. As a first attempt, we focus on short-term electricity consumption in France, which is of strategic importance for energy suppliers and public stakeholders. The complexity of this problem and the many levels of geospatial granularity motivate the use of an ensemble of Gaussian Processes (GPs). Whilst GPs are remarkable predictors, they are computationally expensive to train, which calls for a frugal few-shot learning approach. By taking into account performance on GPs trained on a dataset and designing a random walk on these, we mitigate the training cost of our entire Bayesian decision-making procedure. We introduce our algorithm called \textsc{Domino} (ranDOM walk on gaussIaN prOcesses) and present numerical experiments to support its merits.
Submission Number: 47
Loading