NOVA Framework Discovers Interpretable Models of Driver Behavior from Highway Trajectory Data
Researchers have introduced NOVA, an autonomous symbolic regression framework that derives interpretable mathematical models of car-following and lane-changing behavior directly from raw driving trajectory data. Applied to nearly 4.8 million active driving observations from the NGSIM highway datasets, NOVA outperforms existing symbolic regression baselines on acceleration prediction and surpasses lane-change modeling benchmarks by nearly 30 percentage points in balanced accuracy. The work is significant because it produces compact, human-readable models that generalize across freeway sites without retraining, potentially bridging the gap between black-box machine learning and theory-grounded traffic science.
NOVA is a symbolic regression system built around a deterministic, Rust-powered search engine that evaluates over 10,000 candidate algebraic structures to identify the most predictive and interpretable formulas for driver behavior. Trained and tested on 4,765,788 active driving observations from the widely used NGSIM I-80 and US-101 freeway datasets, the framework discovers a compact two-term acceleration model that achieves an RMSE of 1.376 m/s² (R² = 15.57%) on an intent-forecasting benchmark, improving upon the best recalibrated symbolic regression baseline—SR-LLM from PNAS 2025—by 0.135 m/s² under identical evaluation conditions. A single dominant nonlinear term consistently emerges across eight independent experiments as a robust backbone of human car-following behavior, and a residual-guided extension connects the discovered structure to an established psychophysical theory of collision avoidance. The discovered feature operators transfer zero-shot between the two freeway sites with less than 3 percentage points of R² degradation, suggesting strong generalizability. Extended to lane-change modeling within a multinomial logit framework, NOVA achieves 67.4% balanced accuracy on 502 unseen drivers under strict vehicle-ID holdout, outperforming existing lane-change baselines by 29.8 percentage points on a three-class classification problem. The approach requires minimal behavioral priors, making it broadly applicable to other trajectory datasets and driving environments.
What's missing
The study is a preprint and has not yet undergone peer review. Key limitations and open questions include: the restriction to U.S. freeway data (NGSIM), leaving generalizability to urban, non-U.S., or mixed-traffic environments untested; and whether the psychophysical collision-avoidance link identified in the residual analysis holds under causal or experimental validation rather than correlational discovery.
What different sources said
- arXiv cs.LGCenter
NOVA: Symbolic Regression Discovery of Interpretable Car-Following and Lane-Change Models with Driver Heterogeneity
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.