Hi everyone,
I am working on urban traffic prediction. I want to predict the traffic volume on multiple segments of a road network. The dataset has an averaging interval of 5 min. I want to select the optimal time lags (the horizon of the past traffic volume values) that need to be used to predict the traffic volume in future time steps (15min, 30min, 45min, and 1h). Would anybody help me to find the proper temporal feature selection method?
Any help will be appreciated.