NOVA: Symbolic Regression Discovery of Interpretable Car-Following and Lane-Change Models with Driver Heterogeneity
arXiv:2606.10583v1 Announce Type: new Abstract: We present NOVA, an autonomous symbolic regression framework that identifies interpretable car-following and lane-change structures from raw trajectory data with minimal behavioral priors. Applied to 4,765,788 active driving observations from the NGSIM I-80 and US-101 datasets, NOVA's deterministic Rust-powered search engine evaluates over 10,000 candidate algebraic structures and identifies a compact two-term acceleration model under a forward-shifted rolling-mean prediction target. Evaluated under two complementary preprocessing pipelines, NOVA
The proliferation of high-resolution trajectory data and advancements in symbolic AI techniques enable the automated discovery of complex models without heavy human intervention.
This development allows for the automated generation of highly interpretable and accurate behavioral models for autonomous systems, reducing development costs and increasing safety validation.
The paradigm for developing car-following and lane-change models shifts from expert-driven hypothesis testing to data-driven automated discovery, accelerating progress in autonomous vehicle development.
- · Autonomous vehicle developers
- · Robotics research institutions
- · Logistics and transportation companies
- · Traditional traffic modeling consultants
- · Companies reliant on hand-coded behavioral rules
More robust and safely verifiable autonomous driving systems are developed.
Reduced incidence of traffic accidents, leading to lower insurance costs and improved urban mobility.
Ethical considerations surrounding the explainability and bias of AI-discovered models become paramount as adoption scales.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG