At the forefront of discovery, where cutting-edge scientific questions are tackled, we often don’t have much data. Conversely, successful machine learning (ML) tends to rely on large, high quality ...