A lazy approach to neural numerical planning with control parameters
Publication date
2024
Document type
Conference paper
Author
Organisational unit
Conference
27th European Conference on Artificial Intelligence (ECAI 2024) ; Santiago de Compostela, Spain ; October 19–24, 2024
Publisher
IOS Press
Series or journal
Frontiers in Artificial Intelligence and Applications
ISSN
Periodical volume
392
Book title
ECAI 2024
First page
4262
Last page
4270
Peer-reviewed
✅
Part of the university bibliography
✅
Language
English
Keyword
dtec.bw
Abstract
In this paper, we tackle the problem of planning in complex numerical domains, where actions are indexed by control parameters, and their effects may be described by neural networks. We propose a lazy, hierarchical approach based on two ingredients. First, a Satisfiability Modulo Theory solver looks for an abstract plan where the neural networks in the model are abstracted into uninterpreted functions. Then, we attempt to concretize the abstract plan by querying the neural network to determine the control parameters. If the concretization fails and no valid control parameters could be found, suitable information to refine the abstraction is lifted to the Satisfiability Modulo Theory model. We contrast our work against the state of the art in NN-enriched numerical planning, where the neural network is eagerly and exactly represented as terms in Satisfiability Modulo Theories over nonlinear real arithmetic. Our systematic evaluation on four different planning domains shows that avoiding symbolic reasoning about the neural network not only leads to substantial efficiency improvements, but also enables their integration as black-box models.
Description
This article is published online with Open Access by IOS Press and distributed under the terms of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0).
Version
Published version
Access right on openHSU
Metadata only access