mlsanityData to 14 Oct 2026

Local Repair of Neural Networks Using Optimization

arXiv 2109.14041 ↗First posted 28 Sep 2021cs.LG cs.SY eess.SYcited by 3

Not on the map. Why?

Abstract (arXiv)

In this paper, we propose a framework to repair a pre-trained feed-forward neural network (NN) to satisfy a set of properties. We formulate the properties as a set of predicates that impose constraints on the output of NN over the target input domain. We define the NN repair problem as a Mixed Integer Quadratic Program (MIQP) to adjust the weights of a single layer subject to the given predicates while minimizing the original loss function over the original training domain. We demonstrate the application of our framework in bounding an affine transformation, correcting an erroneous NN in classification, and bounding the inputs of a NN controller.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 16 of its 35 references that are in the map. First posted on arXiv 28 Sep 2021.

Builds on

16 of 35 references.

References · 16

  1. 2019Control Barrier Functions: Theory and Applicationscited by 372
  2. 2017DeepTest: Automated Testing of Deep-Neural-Network-driven Autonomous Carscited by 251
  3. 2019A Survey of Deep Learning Applications to Autonomous Vehicle Controlcited by 106
  4. 2018Formal Verification of Neural Network Controlled Autonomous Systemscited by 41
  5. 2021Partition-based formulations for mixed-integer optimization of trained ReLU neural networkscited by 39
  6. 2017Compositional Falsification of Cyber-Physical Systems with Machine Learning Componentscited by 32
  7. 2020GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learningcited by 21
  8. 2019Requirements-driven Test Generation for Autonomous Vehicles with Machine Learning Componentscited by 18
  9. 2017Deep Predictive Models for Collision Risk Assessment in Autonomous Drivingcited by 15
  10. 2018Semantic Adversarial Deep Learningcited by 15
  11. 2018Counterexample-Guided Data Augmentationcited by 15
  12. 2020Towards Repairing Neural Networks Correctlycited by 14
4 more
  1. 2020Training Neural Network Controllers Using Control Barrier Functions in the Presence of Disturbancescited by 9
  2. 2021Safe-by-Repair: A Convex Optimization Approach for Repairing Unsafe Two-Level Lattice Neural Network Controllerscited by 6
  3. 2019Clone Swarms: Learning to Predict and Control Multi-Robot Systems by Imitationcited by 4
  4. 2021Safe Navigation in Human Occupied Environments Using Sampling and Control Barrier Functionscited by 2

Built on by

Cited by year

Citing papers per year: 2023: 3.

3 papers.

  1. 2023Repairing Deep Neural Networks Based on Behavior Imitationcited by 4
  2. 2023Reachability Analysis of Neural Networks with Uncertain Parameterscited by 0
  3. 2023Causal Repair of Learning-enabled Cyber-physical Systemscited by 0