Clarifying Forecasting Track Challenge

NeurIPS 2023 Citylearn Challenge
#1

I have a few questions

  1. Do we have to use RL models for the forecasting track?
  2. What are the .pth files used for? It says its for “Each building’s temperature dynamics LSTM model parameters”, so do we have to use an LSTM?
  3. When predicting the target, do we have to manually introduce a 48 period lag? Or does the data already account for that?
#2

@ar1 sorry for the late response; here are my answers:

  1. No you don need to use RL models for forecast track.
  2. For the forecast track you don’t have a use for the LSTM models whose parameters are stored in in the .pth files. For the control track, it is used to predict the building temperature based on some provided cooling energy every time step. The controller provides an action for the building’s cooling_device (heat pump), the action is used to calculate the timestep’s cooling demand and then that cooling demand is used as part of the input variables in the LSTM to predict the building’s temperature. You can reference this work for more information about the methodology. You don’t need to use it for anything as it is just the environment that needs it for its inner processes.

@max_langtry please, can you answer the 3rd question?

💬 Feedback & Suggestions
#3

@ar1, my apologies, I’m not sure I quiet understand your 3rd point.
At each time instance your prediction model needs to take a dictionary of observations as its only argument, and return time series of predicted values over the following 48 hours for each prediction variable. The example prediction model provided in the forecasting track starter kit details the required formatting and shows you what a prediction looks like.
Does this answer your question?