Sketching Datasets for Large-Scale Learning (long version)

It is an extended version of https://hal.inria.fr/hal-03350599 (official version published with DOI: https://doi.org/10.1109/MSP.2021.3092574) with additional references and more in-depth discussions on a variety of topics. A python notebook is available at https://gitlab.inria.fr/SketchedLearning/spm-notebookThis article considers "sketched learning," or "compressive learning," an approach to large-scale machine learning where datasets are massively compressed before learning (e.g., clustering, classification, or regression) is performed. In particular, a "sketch" is first constructed by computing carefully chosen nonlinear random features (e.g., random Fourier features) and averaging them over the whole dataset. Parameters are then learned from the sketch, without access to the original dataset. This article surveys the current state-of-the-art in sketched learning, including the main concepts and algorithms, their connections with established signal-processing methods, existing theoretical guarantees-on both information preservation and privacy preservation, and important open problems