dorsal/arxiv
View SchemaBelief Propagation and Bethe approximation for Traffic Prediction
| Authors | Cyril Furtlehner, Jean-Marc Lasgouttes, Arnaud De La Fortelle |
|---|---|
| Categories | |
| ArXiv ID | physics/0703159 |
| URL | https://arxiv.org/abs/physics/0703159 |
Abstract
We define and study an inference algorithm based on "belief propagation" (BP) and the Bethe approximation. The idea is to encode into a graph an a priori information composed of correlations or marginal probabilities of variables, and to use a message passing procedure to estimate the actual state from some extra real-time information. This method is originally designed for traffic prediction and is particularly suitable in settings where the only information available is floating car data. We propose a discretized traffic description, based on the Ising model of statistical physics, in order to both reconstruct and predict the traffic in real time. General properties of BP are addressed in this context. In particular, a detailed study of stability is proposed with respect to the a priori data and the graph topology. The behavior of the algorithm is illustrated by numerical studies on a simple traffic toy model. How this approach can be generalized to encode superposition of many traffic patterns is discussed.
{
"annotation_id": "6bf09858-63b7-4fa4-8551-39438374f09a",
"date_created": "2026-03-02T18:01:18.548000Z",
"date_modified": "2026-03-02T18:01:18.548000Z",
"file_hash": "0ae9d206bf379e0477a8b457b7e9181cf0201707eac0e0843cfba77cf256dc86",
"private": false,
"record": {
"abstract": "We define and study an inference algorithm based on \"belief propagation\" (BP)\nand the Bethe approximation. The idea is to encode into a graph an a priori\ninformation composed of correlations or marginal probabilities of variables,\nand to use a message passing procedure to estimate the actual state from some\nextra real-time information. This method is originally designed for traffic\nprediction and is particularly suitable in settings where the only information\navailable is floating car data. We propose a discretized traffic description,\nbased on the Ising model of statistical physics, in order to both reconstruct\nand predict the traffic in real time. General properties of BP are addressed in\nthis context. In particular, a detailed study of stability is proposed with\nrespect to the a priori data and the graph topology. The behavior of the\nalgorithm is illustrated by numerical studies on a simple traffic toy model.\nHow this approach can be generalized to encode superposition of many traffic\npatterns is discussed.",
"arxiv_id": "physics/0703159",
"authors": [
"Cyril Furtlehner",
"Jean-Marc Lasgouttes",
"Arnaud De La Fortelle"
],
"categories": [
"physics.soc-ph",
"math.PR"
],
"title": "Belief Propagation and Bethe approximation for Traffic Prediction",
"url": "https://arxiv.org/abs/physics/0703159"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "498b82ea-78bc-4f20-9baa-9d22bac64527",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}