We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Guaranteed Conformance of Neurosymbolic Models to Natural Constraints

Abstract: Deep neural networks have emerged as the workhorse for a large section of robotics and control applications, especially as models for dynamical systems. Such data-driven models are in turn used for designing and verifying autonomous systems. This is particularly useful in modeling medical systems where data can be leveraged to individualize treatment. In safety-critical applications, it is important that the data-driven model is conformant to established knowledge from the natural sciences. Such knowledge is often available or can often be distilled into a (possibly black-box) model $M$. For instance, the unicycle model (which encodes Newton's laws) for an F1 racing car. In this light, we consider the following problem - given a model $M$ and state transition dataset, we wish to best approximate the system model while being bounded distance away from $M$. We propose a method to guarantee this conformance. Our first step is to distill the dataset into few representative samples called memories, using the idea of a growing neural gas. Next, using these memories we partition the state space into disjoint subsets and compute bounds that should be respected by the neural network, when the input is drawn from a particular subset. This serves as a symbolic wrapper for guaranteed conformance. We argue theoretically that this only leads to bounded increase in approximation error; which can be controlled by increasing the number of memories. We experimentally show that on three case studies (Car Model, Drones, and Artificial Pancreas), our constrained neurosymbolic models conform to specified $M$ models (each encoding various constraints) with order-of-magnitude improvements compared to the augmented Lagrangian and vanilla training methods. Our code can be found at this https URL
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO)
Cite as: arXiv:2212.01346 [cs.LG]
  (or arXiv:2212.01346v6 [cs.LG] for this version)

Submission history

From: Kaustubh Sridhar [view email]
[v1] Fri, 2 Dec 2022 18:03:37 GMT (4677kb,D)
[v2] Thu, 8 Dec 2022 11:49:53 GMT (4802kb,D)
[v3] Fri, 9 Dec 2022 23:58:50 GMT (4802kb,D)
[v4] Wed, 8 Feb 2023 07:01:33 GMT (4802kb,D)
[v5] Sun, 12 Feb 2023 05:37:19 GMT (4802kb,D)
[v6] Thu, 9 Mar 2023 23:17:08 GMT (4212kb,D)

Link back to: arXiv, form interface, contact.