Dataset Distillation with Infinitely Wide Convolutional Networks
This work addresses the problem of training efficiency and feature extraction for machine learning practitioners by enabling highly performant models with drastically reduced data, representing a strong incremental advance in dataset distillation.
The paper tackles dataset compression by distilling large datasets into much smaller ones using a distributed kernel meta-learning framework with infinitely wide convolutional networks, achieving over 65% test accuracy on CIFAR-10 with only 10 datapoints, a significant improvement from the previous best of 40%.
The effectiveness of machine learning algorithms arises from being able to extract useful features from large amounts of data. As model and dataset sizes increase, dataset distillation methods that compress large datasets into significantly smaller yet highly performant ones will become valuable in terms of training efficiency and useful feature extraction. To that end, we apply a novel distributed kernel based meta-learning framework to achieve state-of-the-art results for dataset distillation using infinitely wide convolutional neural networks. For instance, using only 10 datapoints (0.02% of original dataset), we obtain over 65% test accuracy on CIFAR-10 image classification task, a dramatic improvement over the previous best test accuracy of 40%. Our state-of-the-art results extend across many other settings for MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100, and SVHN. Furthermore, we perform some preliminary analyses of our distilled datasets to shed light on how they differ from naturally occurring data.