DrivAerML Dataset Files

The DrivAerML dataset consists of 500 parametrically varied geometries. However, only 484 of these geometries are usable for training and validation. Since the publication does not recommend any training-validation split, we propose a split in this repository to enable a fair and consistent benchmark.

The dataset is split as 90% for training and 10% for validation. For selecting the validation set, the entire dataset is sorted by drag force. Then 10% of the validation set is chosen from the top of the sorted dataset, and another 10% of the validation set is chosen from the bottom of the sorted dataset. The remaining 80% of the validation set is chosen randomly from the remaining dataset.

This allows us to include some out-of-distribution data in the validation set, which is important for the generalizability of the models.

The figure below shows the distribution of the training-validation sets.