Some clarifications to my question:

  • The data I have available for use is: (a) historic data of features and ground truth on 60-minute interval, (b) real-time data of features on 60-minute interval, (c) real-time data of ground truth on 15-minute interval.
  • The use case is in energy yield forecasting, where we want to improve short-term, intra-hour predictions for photovoltaic energy production (yield). We have historic and real-time measurement data from photovoltaic systems, as well as dumps of historic weather forecasts and actual weather forecasts for the next 1-36 hours.
  • So far, we’re using regression-based machine learning models and historic data (case a)) to create (intra-day) forecasts on a 60-min interval for a 12–36h time horizon.
  • We have the possibility to include real-time measurements (via web-APIs) and would love to include these to forecast the energy production for just the next 15 minutes. I imagine using our prediction model for intra-day forecasts and combine it with real-time measurements (e.g. of the same day or just the previous 15 minutes) to predict the next 15 minutes. Could this be done with some kind of sliding time window?
  • The target forecasting interval is 15-minutes; since some of our data are available on a 60-min interval only, I am assuming this requires interpolation.
  • Since photovoltaic yields are especially impacted by the weather, we have daily and yearly seasonality effects.

I've already identified the following approaches through web searches and reading related work. However, I don't yet understand how they consider/include the real-time data:

  • Recurrent Neural Networks (LSTM or GRU)
  • Prophet from Meta (but it seems to not be actively developed anymore)
  • Reinforcement Learning (but we're not modifying the environment)
  • $\begingroup$ The answer is: whichever works for you (your data)! $\endgroup$ Commented Dec 20, 2023 at 14:33
  • $\begingroup$ Welcome to CV! What is your goal here? The question in your title is clear, but perhaps you can explain why you think these techniques may help you answer that question. $\endgroup$ Commented Dec 20, 2023 at 15:25
  • $\begingroup$ ARIMA/SARIMA do not use features (although you can run a regression with ARIMA errors). You could look at the GEFCom2017 competition and see what people used there to leverage predictors. $\endgroup$ Commented Dec 20, 2023 at 20:01
  • $\begingroup$ Thanks for your comments. I've tried to clarify the goal a bit in my question. I've identified some possible techniques, but am unsure if and how these allow to include real-time data to improve the prediction of just the next 15-minute energy production. $\endgroup$
    – casaout
    Commented Dec 22, 2023 at 7:46

1 Answer 1


In my case, I have managed to succesfully implement forecasting systems like the one you are describing (with a rolling window approach) in environments with considerably seasonal multivariate time series (you can check the results I obtained here https://ieeexplore.ieee.org/document/10205609).

ARIMA and SARIMAX might require more manual re-estimation when data patterns change significantly.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.