A new measure for overfitting and its implications for backdooring of deep learning

06/11/2020
by   Kathrin Grosse, et al.
0

Overfitting describes the phenomenon that a machine learning model fits the given data instead of learning the underlying distribution. Existing approaches are computationally expensive, require large amounts of labeled data, consider overfitting global phenomenon, and often compute a single measurement. Instead, we propose a local measurement around a small number of unlabeled test points to obtain features of overfitting. Our extensive evaluation shows that the measure can reflect the model's different fit of training and test data, identify changes of the fit during training, and even suggest different fit among classes. We further apply our method to verify if backdoors rely on overfitting, a common claim in security of deep learning. Instead, we find that backdoors rely on underfitting. Our findings also provide evidence that even unbackdoored neural networks contain patterns similar to backdoors that are reliably classified as one class.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
06/26/2019

Benign Overfitting in Linear Regression

The phenomenon of benign overfitting is one of the key mysteries uncover...
research
07/14/2022

Benign, Tempered, or Catastrophic: A Taxonomy of Overfitting

The practical success of overparameterized neural networks has motivated...
research
06/01/2022

Realistic Deep Learning May Not Fit Benignly

Studies on benign overfitting provide insights for the success of overpa...
research
05/09/2023

Testing for Overfitting

High complexity models are notorious in machine learning for overfitting...
research
05/24/2023

From Tempered to Benign Overfitting in ReLU Neural Networks

Overparameterized neural networks (NNs) are observed to generalize well ...
research
03/06/2019

Detecting Overfitting via Adversarial Examples

The repeated reuse of test sets in popular benchmark problems raises dou...
research
04/28/2021

Risk Bounds for Over-parameterized Maximum Margin Classification on Sub-Gaussian Mixtures

Modern machine learning systems such as deep neural networks are often h...

Please sign up or login with your details

Forgot password? Click here to reset