Michael Unser: Splines and Machine Learning: From classical RKHS methods to DNN (MLSP 2020 keynote)

Splines and Machine Learning: From classical RKHS methods to deep neural networks.

Speaker: Michael Unser, École polytechnique fédérale de Lausanne (EPFL).

Abstract: Supervised learning is a fundamentally ill-posed problem. In practice, this indetermination is dealt with by imposing constraints on the solution; these are either implicit, as in neural networks, or explicit via the use of a regularization functional. In this talk, I present a unifying perspective that revolves around a new representer theorem that characterizes the solution of a broad class of functional optimization problems. I then use this theorem to derive the most prominent classical algorithms — e.g., kernel-based techniques and smoothing splines — as well as their “sparse” counterparts. This leads to the identification of sparse adaptive splines, which have some remarkable properties.

I then show how the latter can be integrated in conventional neural architectures to yield high-dimensional adaptive linear splines. Finally, I recover deep neural nets with ReLU activations as a particular case.

M. Unser, “A unifying representer theorem for inverse problems and machine learning,” Foundations of Computational Mathematics, in press, 2020.
M. Unser, “A representer theorem for deep neural networks,” Journal of Machine Learning Research, vol. 20(110), pp. 1–30, 2019.

Presented at IEEE International Workshop on Machine Learning for Signal Processing, Monday, September 21, 2020.