A Network Screening Method for Short-Term Safety Performance
Network screening is crucial to transportation safety management, enabling practitioners to identify high-risk locations and implement countermeasures to reduce crashes. Traditional methods focus on long-term crash trends using multi-year data. Short-term network screening—analyzing crash patterns over weeks to months—remains underexplored. This gap may lead to missed opportunities by overlooking locations with short-term crash hotspots that could be mitigated through enforcement or other low-cost interventions. Additionally, distinguishing between long-term and short-term hotspots is essential for effective safety mitigation. To address this gap, this study developed a short-term network screening framework integrating observed crashes and predictive modeling to identify crash hotspots. Using weekly crash data from the Iowa Interstate network, the study employs a binary logistic regression model to estimate predicted crash probabilities. Then, historically observed crash data is combined with the predicted crash probabilities to estimate the final crash likelihood of a particular interstate segment. The framework was then validated with a 12-week validation period data to identify the performance of the network screening process. Despite using a highly skewed dataset where 98% of records indicated no crashes, the proposed framework demonstrated effective performance compared to the validation data, proving its usefulness for highly skewed datasets. Results also show that the proposed framework successfully identifies the short-term crash hotspots, with ranking correlation values larger than 0.8 across different analysis periods, further indicating the framework's reliability and consistency.
Code (0)
등록된 구현이 없습니다.
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Machine learning approach for early detection of autism by combining questionnaire and home video screening
Existing screening tools for early detection of autism are expensive, cumbersome, time-intensive, and sometimes fall short in predictive value. In this work, we apply Machine Learning (ML) to gold standard clinical data …
BIG-bench Machine LearningFeature Engineeringfeature selectionVetClaw: An Edge-Cloud Multimodal Agentic System for Veterinary Disease Screening
We present VetClaw, an edge-cloud multimodal agentic system for early veterinary disease screening. VetClaw uses a camera module as an edge sensing device and sends captured images, together with optional symptom descrip…
Image ClassificationStable safe screening and structured dictionaries for faster L1 regularization
In this paper, we propose a way to combine two acceleration techniques for the $\ell\_{1}$-regularized least squares problem: safe screening tests, which allow to eliminate useless dictionary atoms; and the use of fast s…
SafeScreen: A Safety-First Screening Framework for Personalized Video Retrieval for Vulnerable Users
Open-domain video platforms offer rich, personalized content that could support health, caregiving, and educational applications, but their engagement-optimized recommendation algorithms can expose vulnerable users to in…
Question GenerationVideo RetrievalA new Time-decay Radiomics Integrated Network (TRINet) for short-term breast cancer risk prediction
To facilitate early detection of breast cancer, there is a need to develop short-term risk prediction schemes that can prescribe personalized/individualized screening mammography regimens for women. In this study, we pro…
Continual LearningMultiple Instance Learning