Probabilistic Lipschitz analysis of neural networks

DOI10.1007/978-3-030-65474-0_13MaRDI QIDQ2233541zbMATH OpenOpenAlexFDO

Authors Ravi Mangal, Kartik Sarangmath, Aditya V. Nori, Alessandro Orso

Publication date 18 October 2021

Full work available at URL https://doi.org/10.1007/978-3-030-65474-0_13

Mathematics Subject Classification ID

Theory of programming languages (68N15) Mathematical aspects of software engineering (specification, verification, metrics, requirements, etc.) (68N30) Semantics in the theory of computing (68Q55) Networks and circuits as models of computation; circuit complexity (68Q06)

Recommendations

• Probabilistic robustness estimates for feed-forward neural networks
• Enhancing robustness verification for deep neural networks via symbolic propagation
• CLIP: cheap Lipschitz training of neural networks
• Robustness and generalization
• Lipschitz properties for deep convolutional networks

Cites work

• A geometric inequality and the complexity of computing volume
• A random polynomial-time algorithm for approximating the volume of convex bodies
• APLicative Programming with Naperian Functors
• An array-oriented language with static rank polymorphism
• An assertion-based program logic for probabilistic programs
• Computing the volume is difficult
• Gaussian Cooling and $O^*(n^3)$ Algorithms for Volume and Gaussian Volume
• Linear-invariant generation for probabilistic programs: automated support for proof-based methods
• On the Complexity of Computing the Volume of a Polyhedron
• Probabilistic abstract interpretation
• Proving the Correctness of Multiprocess Programs
• Reluplex: an efficient SMT solver for verifying deep neural networks
• Simple relational correctness proofs for static analyses and program transformations
• Synthesizing Probabilistic Invariants via Doob’s Decomposition

Cited in

(9)

• CLIP: cheap Lipschitz training of neural networks
• ABBA neural networks: coping with positivity, expressivity, and robustness
• Regularisation of neural networks by enforcing Lipschitz continuity
• Probabilistic robustness estimates for feed-forward neural networks
• \textsf{CLEVEREST}: accelerating CEGAR-based neural network verification via adversarial attacks
• Lipschitz Certificates for Layered Network Structures Driven by Averaged Activation Operators
• On Lipschitz Bounds of General Convolutional Neural Networks
• Robustness analysis of continuous-depth models with Lagrangian techniques
• LinSyn: synthesizing tight linear bounds for arbitrary neural network activation functions

This page was built for publication: Probabilistic Lipschitz analysis of neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2233541)