Our paper, Making Classical Machine Learning Pipelines Differentiable: A Neural Translation Approach, was accepted to Systems for ML Workshop at NIPS 2018! This is joint work with Microsoft. Congrats, Gyeong-In!