Skip to content

Evaluating Robustness of Neural Networks [Virtual]

Photo of Ryan Chesler
Hosted By
Ryan C. and Ted K.
Evaluating Robustness of Neural Networks [Virtual]

Details

Evaluating Robustness of Neural Networks
by Lily Weng

Abstract:
The robustness of neural networks to adversarial examples has received great attention due to security implications. Despite various attack approaches to crafting visually imperceptible adversarial examples, little has been developed towards a comprehensive measure of robustness. In this talk, Weng will introduce several robustness quantification frameworks for deep neural networks against both adversarial and non-adversarial input perturbations, including the first robustness score CLEVER, efficient certification algorithms Fast-Lin, CROWN, CNN-Cert, and probabilistic robustness verification algorithm PROVEN. The proposed approaches are computationally efficient and provide a good quality of robustness estimate/certificate as demonstrated by extensive experiments on MNIST, CIFAR and ImageNet.

Bio:
Lily Weng is an Assistant Professor in the Halicioglu Data Science Institute at UC San Diego with an affiliation to the CSE department. She has broad research interest in the intersection of machine learning, optimization and reinforcement learning, with applications in cybersecurity and healthcare. Her vision is to make the next generation AI systems and deep learning algorithms more robust, reliable, trustworthy and safer. She has worked on developing efficient algorithms as well as theoretical analysis to quantify robustness of deep neural networks. She received her PhD in Electrical Engineering and Computer Sciences (EECS) from MIT in 2020, and her Bachelor and Master degree both in Electrical Engineering at National Taiwan University in 2011 and 2013.

=================
Agenda (Pacific Daylight Time, UTC -07)

  • 5:30 - 5:40 pm -- Gathering and introductions
  • 5:40 - 6:30 pm -- Talk
  • 6:30 - 7:00 pm -- Q & A, discussion

Links to slides and videos of meetup presentations are available on the SDML GitHub repo https://github.com/SanDiegoMachineLearning/talks

=================
Questions?

Join our slack channel or leave a comment below if you have any questions about the group or need clarification on anything.
https://join.slack.com/t/sdmachinelearning/shared_invite/zt-6b0ojqdz-9bG7tyJMddVHZ3Zm9IajJA

Photo of San Diego Machine Learning group
San Diego Machine Learning
See more events