My team have read the rules for this competition and the phrase "You may use only the datasets provided for this competition" is not clear for us. Does it mean that we MUST use only the data provided? Or does it mean that we may use only the data provided by choice?
You must only use the datasets for this challenge, you cannot add external data.
Follow-up question: Are external, publicly available data sets for pretraining a model allowed? Asking as from my point of view adding this kind of data wouldn't change the nature of the challenge (i.e. the used input features for prediction) but may allow to benefit from transfer learning