Learning Constraints Directly from Network Data

• URL: http://arxiv.org/abs/2506.23964v1
• Date: Mon, 30 Jun 2025 15:36:22 GMT
• Title: Learning Constraints Directly from Network Data
• Authors: Hongyu Hè, Minhao Jin, Maria Apostolaki,
• Abstract summary: Rule extraction can improve quality of synthetic data, reduce brittleness of machine learning models, and improve semantic understanding of network measurements.<n>This paper introduces NetNomos, which learns propositional logic constraints directly from raw network measurements.<n>Our evaluations show that NetNomos learns all benchmark rules, including those associated with as little as 0.01% of data points, in under three hours.
• Score: 0.34137115855910755
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Network data conforms to a wide range of rules that arise from protocols, design principles, and deployment decisions (e.g., a packet's queuing delay must be less than its end-to-end delay). Formalizing such rules as logic constraints can (i) improve the quality of synthetic data, (ii) reduce the brittleness of machine learning (ML) models, and (iii) improve semantic understanding of network measurements. However, these benefits remain out of reach if rule extraction is manual or solely reliant on ML, as both approaches yield incomplete, unreliable, and/or inaccurate rules. This paper formulates rule extraction as a constraint modeling problem and introduces NetNomos that learns propositional logic constraints directly from raw network measurements. Constraint modeling in this domain is uniquely challenging due to the scale of the data, the inherent learning complexity and passive environment, and the lack of ground truth supervision. NetNomos addresses these challenges via a lattice-based search structured by constraint specificity and succinctness. Our approach reduces learning complexity from superquadratic to logarithmic and enables efficient traversal in combinatorial search space. Our evaluations on diverse network datasets show that NetNomos learns all benchmark rules, including those associated with as little as 0.01% of data points, in under three hours. In contrast, baseline methods discover less than 25% of the rules and require several days to run. Through three case studies, we show that: NetNomos (i) finds rule violations in the outputs of all seven synthetic traffic generators, hence can be used to assess and guide their generation process; (ii) detects semantic differences in traffic, hence can be used for anomaly detection; and (iii) automatically finds rules used for telemetry imputation, hence can support monitoring through inference.

Related papers

• A General Framework for Property-Driven Machine Learning [1.0485739694839669]
  We show that adversarial training can be used to improve robustness to small perturbations within $epsilon$-cubes.<n> domains other than computer vision may require more flexible input region specifications via generalised hyper-rectangles.<n>In this paper, we investigate how these two complementary approaches can be unified within a single framework for property-driven machine learning.
  arXiv  Detail & Related papers  (2025-05-01T11:33:38Z)
• Using Petri Nets as an Integrated Constraint Mechanism for Reinforcement Learning Tasks [3.105112058253643]
  Lack of trust in algorithms is usually an issue when using Reinforcement Learning (RL) agents for control in real-world domains. We propose an approach that uses Petri nets (PNs) with three main advantages over typical RL approaches.
  arXiv  Detail & Related papers  (2024-07-05T13:04:06Z)
• Decentralized Learning Strategies for Estimation Error Minimization with Graph Neural Networks [94.2860766709971]
  We address the challenge of sampling and remote estimation for autoregressive Markovian processes in a wireless network with statistically-identical agents.<n>Our goal is to minimize time-average estimation error and/or age of information with decentralized scalable sampling and transmission policies.
  arXiv  Detail & Related papers  (2024-04-04T06:24:11Z)
• Optimization Guarantees of Unfolded ISTA and ADMM Networks With Smooth Soft-Thresholding [57.71603937699949]
  We study optimization guarantees, i.e., achieving near-zero training loss with the increase in the number of learning epochs. We show that the threshold on the number of training samples increases with the increase in the network width.
  arXiv  Detail & Related papers  (2023-09-12T13:03:47Z)
• Robust Online Learning over Networks [1.0249620437941]
  This work specifically targets some prevalent challenges inherent to distributed learning. We apply the Distributed Operator Theoretical (DOT) version of the Alternating Direction Method of Multipliers (ADMM) We prove that if the DOT-ADMM operator is metric subregular, then it converges with a linear rate for a class of convex learning problems.
  arXiv  Detail & Related papers  (2023-09-01T15:18:05Z)
• Networked Communication for Decentralised Agents in Mean-Field Games [59.01527054553122]
  We introduce networked communication to the mean-field game framework.<n>We prove that our architecture has sample guarantees bounded between those of the centralised- and independent-learning cases.<n>We show that our networked approach has significant advantages over both alternatives in terms of robustness to update failures and to changes in population size.
  arXiv  Detail & Related papers  (2023-06-05T10:45:39Z)
• Synergy and Symmetry in Deep Learning: Interactions between the Data, Model, and Inference Algorithm [33.59320315666675]
  We study the triplet (D, M, I) as an integrated system and identify important synergies that help mitigate the curse of dimensionality. We find that learning is most efficient when these symmetries are compatible with those of the data distribution.
  arXiv  Detail & Related papers  (2022-07-11T04:08:21Z)
• Interpolation-based Contrastive Learning for Few-Label Semi-Supervised Learning [43.51182049644767]
  Semi-supervised learning (SSL) has long been proved to be an effective technique to construct powerful models with limited labels. Regularization-based methods which force the perturbed samples to have similar predictions with the original ones have attracted much attention. We propose a novel contrastive loss to guide the embedding of the learned network to change linearly between samples.
  arXiv  Detail & Related papers  (2022-02-24T06:00:05Z)
• Characterizing possible failure modes in physics-informed neural networks [55.83255669840384]
  Recent work in scientific machine learning has developed so-called physics-informed neural network (PINN) models. We demonstrate that, while existing PINN methodologies can learn good models for relatively trivial problems, they can easily fail to learn relevant physical phenomena even for simple PDEs. We show that these possible failure modes are not due to the lack of expressivity in the NN architecture, but that the PINN's setup makes the loss landscape very hard to optimize.
  arXiv  Detail & Related papers  (2021-09-02T16:06:45Z)
• SignalNet: A Low Resolution Sinusoid Decomposition and Estimation Network [79.04274563889548]
  We propose SignalNet, a neural network architecture that detects the number of sinusoids and estimates their parameters from quantized in-phase and quadrature samples. We introduce a worst-case learning threshold for comparing the results of our network relative to the underlying data distributions. In simulation, we find that our algorithm is always able to surpass the threshold for three-bit data but often cannot exceed the threshold for one-bit data.
  arXiv  Detail & Related papers  (2021-06-10T04:21:20Z)
• Reinforcement Learning with External Knowledge by using Logical Neural Networks [67.46162586940905]
  A recent neuro-symbolic framework called the Logical Neural Networks (LNNs) can simultaneously provide key-properties of both neural networks and symbolic logic. We propose an integrated method that enables model-free reinforcement learning from external knowledge sources.
  arXiv  Detail & Related papers  (2021-03-03T12:34:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.