Towards Robust, Locally Linear Deep Networks

Deep networks realize complex mappings that are often understood by their locally linear behavior at or around points of interest. For example, we use the derivative of the mapping with respect to its inputs for sensitivity analysis, or to explain (obtain coordinate relevance for) a prediction. One key challenge is that such derivatives are themselves inherently unstable. In this paper, we propose a new learning problem to encourage deep networks to have stable derivatives over larger regions. While the problem is challenging in general, we focus on networks with piecewise linear activation functions. Our algorithm consists of an inference step that identifies a region around a point where linear approximation is provably stable, and an optimization step to expand such regions. We propose a novel relaxation to scale the algorithm to realistic models. We illustrate our method with residual and recurrent networks on image and sequence datasets.

[1]  Luca Antiga,et al.  Automatic differentiation in PyTorch , 2017 .

[2]  Li Fei-Fei,et al.  ImageNet: A large-scale hierarchical image database , 2009, CVPR.

[3]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[4]  Matthias Hein,et al.  Provable Robustness of ReLU networks via Maximization of Linear Regions , 2018, AISTATS.

[5]  Christian Tjandraatmadja,et al.  Bounding and Counting Linear Regions of Deep Neural Networks , 2017, ICML.

[6]  Aleksander Madry,et al.  Towards Deep Learning Models Resistant to Adversarial Attacks , 2017, ICLR.

[7]  Jonathon Shlens,et al.  Explaining and Harnessing Adversarial Examples , 2014, ICLR.

[8]  Motoaki Kawanabe,et al.  How to Explain Individual Classification Decisions , 2009, J. Mach. Learn. Res..

[9]  Aaron C. Courville,et al.  Improved Training of Wasserstein GANs , 2017, NIPS.

[10]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[11]  Tommi S. Jaakkola,et al.  On the Robustness of Interpretability Methods , 2018, ArXiv.

[12]  Darrell Whitley,et al.  A genetic algorithm tutorial , 1994, Statistics and Computing.

[13]  Mykel J. Kochenderfer,et al.  Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks , 2017, CAV.

[14]  Qiang Ye,et al.  Orthogonal Recurrent Neural Networks with Scaled Cayley Transform , 2017, ICML.

[15]  Percy Liang,et al.  Understanding Black-box Predictions via Influence Functions , 2017, ICML.

[16]  Martin Wattenberg,et al.  SmoothGrad: removing noise by adding noise , 2017, ArXiv.

[17]  Andrew L. Maas.  Rectifier Nonlinearities Improve Neural Network Acoustic Models , 2013 .

[18]  Changshui Zhang,et al.  Deep Defense: Training DNNs with Improved Adversarial Robustness , 2018, NeurIPS.

[19]  J. Zico Kolter,et al.  Provable defenses against adversarial examples via the convex outer adversarial polytope , 2017, ICML.

[20]  Yoshua Bengio,et al.  Unitary Evolution Recurrent Neural Networks , 2015, ICML.

[21]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Sanjiv Kumar,et al.  On the Convergence of Adam and Beyond , 2018 .

[23]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[24]  J. Zico Kolter,et al.  Scaling provable adversarial defenses , 2018, NeurIPS.

[25]  Alessio Lomuscio,et al.  An approach to reachability analysis for feed-forward ReLU neural networks , 2017, ArXiv.

[26]  Hossein Mobahi,et al.  Large Margin Deep Networks for Classification , 2018, NeurIPS.

[27]  Tommi S. Jaakkola,et al.  Towards Robust Interpretability with Self-Explaining Neural Networks , 2018, NeurIPS.

[28]  John C. Duchi,et al.  Certifying Some Distributional Robustness with Principled Adversarial Training , 2017, ICLR.

[29]  Marc G. Bellemare,et al.  The Cramer Distance as a Solution to Biased Wasserstein Gradients , 2017, ArXiv.

[30]  Samy Bengio,et al.  Understanding deep learning requires rethinking generalization , 2016, ICLR.

[31]  Thomas Brox,et al.  Striving for Simplicity: The All Convolutional Net , 2014, ICLR.

[32]  Matthew Mirman,et al.  Differentiable Abstract Interpretation for Provably Robust Neural Networks , 2018, ICML.

[33]  Inderjit S. Dhillon,et al.  Towards Fast Computation of Certified Robustness for ReLU Networks , 2018, ICML.

[34]  V. Vapnik.  Estimation of Dependences Based on Empirical Data , 2006 .

[35]  Matteo Fischetti,et al.  Deep Neural Networks as 0-1 Mixed Integer Linear Programs: A Feasibility Study , 2017, ArXiv.

[36]  Yoshua Bengio,et al.  A Generative Process for sampling Contractive Auto-Encoders , 2012, ICML 2012.

[37]  Ayhan Demiriz,et al.  Semi-Supervised Support Vector Machines , 1998, NIPS.

[38]  Richard Baraniuk,et al.  Mad Max: Affine Spline Insights Into Deep Learning , 2018, Proceedings of the IEEE.

[39]  Abubakar Abid,et al.  Interpretation of Neural Networks is Fragile , 2017, AAAI.

[40]  Surya Ganguli,et al.  On the Expressive Power of Deep Neural Networks , 2016, ICML.

[41]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[42]  Ananthram Swami,et al.  The Limitations of Deep Learning in Adversarial Settings , 2015, 2016 IEEE European Symposium on Security and Privacy (EuroS&P).

[43]  Chih-Hong Cheng,et al.  Maximum Resilience of Artificial Neural Networks , 2017, ATVA.

[44]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[45]  Yoshua Bengio,et al.  Deep Sparse Rectifier Neural Networks , 2011, AISTATS.

[46]  Razvan Pascanu,et al.  On the Number of Linear Regions of Deep Neural Networks , 2014, NIPS.

[47]  Dilin Wang,et al.  Learning to Draw Samples: With Application to Amortized MLE for Generative Adversarial Learning , 2016, ArXiv.

[48]  Ankur Taly,et al.  Axiomatic Attribution for Deep Networks , 2017, ICML.

[49]  G. Griffin,et al.  Caltech-256 Object Category Dataset , 2007 .