Distribution Density, Tails, and Outliers in Machine Learning: Metrics and Applications

10/29/2019 ∙ by Nicholas Carlini, et al. ∙ 24

We develop techniques to quantify the degree to which a given (training or testing) example is an outlier in the underlying distribution. We evaluate five methods to score examples in a dataset by how well-represented the examples are, for different plausible definitions of "well-represented", and apply these to four common datasets: MNIST, Fashion-MNIST, CIFAR-10, and ImageNet. Despite being independent approaches, we find all five are highly correlated, suggesting that the notion of being well-represented can be quantified. Among other uses, we find these methods can be combined to identify (a) prototypical examples (that match human expectations); (b) memorized training examples; and, (c) uncommon submodes of the dataset. Further, we show how we can utilize our metrics to determine an improved ordering for curriculum learning, and impact adversarial robustness. We release all metric values on training and test sets we studied.



There are no comments yet.


page 18

page 19

page 20

page 24

page 25

page 33

page 40

page 42

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.