Predictive Uncertainty in Short-Term PV Forecasting under Missing Data: A Multiple Imputation Approach
2026-03-16 • Machine Learning
Machine Learning
AI summaryⓘ
The authors address the common problem of missing data in solar power (photovoltaic) measurements, which can cause uncertainty in forecasts. They created a method that fills in missing data several times to capture uncertainty and combines these results to improve predictions. This method can work with any machine learning model. Their tests show that ignoring the uncertainty from missing data makes the forecast too confident, while their approach gives better calibrated prediction ranges without losing accuracy. Overall, they highlight the need to consider data gaps when predicting solar power output.
photovoltaic powermissing datauncertainty propagationmultiple imputationRubin's ruleshort-term forecastingprediction intervalsmachine learninginterval calibration
Authors
Parastoo Pashmchi, Jérôme Benoit, Motonobu Kanagawa
Abstract
Missing values are common in photovoltaic (PV) power data, yet the uncertainty they induce is not propagated into predictive distributions. We develop a framework that incorporates missing-data uncertainty into short-term PV forecasting by combining stochastic multiple imputation with Rubin's rule. The approach is model-agnostic and can be integrated with standard machine-learning predictors. Empirical results show that ignoring missing-data uncertainty leads to overly narrow prediction intervals. Accounting for this uncertainty improves interval calibration while maintaining comparable point prediction accuracy. These results demonstrate the importance of propagating imputation uncertainty in data-driven PV forecasting.