Application of Conditional Deep Generative Networks (CGAN) in empirical bayes estimation of road crash risk and identifying crash hotspots
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The conditional generative adversarial network (CGAN) is used in this paper for empirical Bayes (EB) analysis of road crash hotspots. EB is a well-known method for estimating the expected crash frequency of sites (e.g. road segments, intersections) and then prioritising these sites to identify a subset of high priority sites (e.g. hotspots) for additional safety audits/improvements. In contrast to the conventional EB approach, which employs a statistical model such as the negative binomial model (NB-EB) to model crash frequency data, the recently developed CGAN-EB approach uses a conditional generative adversarial network, a form of deep neural network, that can model any form of distributions of the crash frequency data. Previous research has shown that the CGAN-EB performs as well as or better than NB-EB, however that work considered only a small range of crash data characteristics and did not examine the spatial and temporal transferability. In this paper a series of simulation experiments are devised and carried out to assess the CGAN-EB performance across a wide range of conditions and compares it to the NB-EB. The simulation results show that CGAN-EB performs as well as NB-EB when conditions favor the NB-EB model (i.e. data conform to the assumptions of the NB model) and outperforms NB-EB in experiments reflecting conditions frequently encountered in practice (i.e. low sample mean crash rates, and when crash frequency does not follow a log-linear relationship with covariates). Also, temporal and spatial transferability of both approaches were evaluated using field data and both CGAN-EB and NB-EB approaches were found to have similar performance.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it