 |  |  |  |
Using Time-Series Analysis to Precisely Identify and Rank Road Hotspots
Principal Investigator:
Alexander Lee, University of Arizona
Co-Investigators:
- Wei Lin, University of Arizona
Summary:
Over the past decades, many ranking methods have been proposed. However, results vary from method to method, and one of the issues behind ranking is the element of subjectivity. One approach to resolve these issues is the use of combined models. One of the combined models is the Enhanced Empirical Bayesian (EB) method that incorporates the use of the similarity measure based on the Proportion Disc...
Over the past decades, many ranking methods have been proposed. However, results vary from method to method, and one of the issues behind ranking is the element of subjectivity. One approach to resolve these issues is the use of combined models. One of the combined models is the Enhanced Empirical Bayesian (EB) method that incorporates the use of the similarity measure based on the Proportion Discordance Ratio (PDR). This model is developed to assess and objectively quantify similarity among road segments based on crash patterns, each of which contains a unique combination of selected crash-related features. The goal of this project is to identify a group of similar road segments for the estimation of road segment safety levels and to identify and rank road hotspots for a particular highway at certain hours of the day and days of the week. Based on this assessment, USDOT can find the root cause of the high risk of crash occurrences, and recommendations can be effectively made on a case-by-case basis to reduce such risk. This project addresses the themes of “Developing Data, Models, and Tools” and “Improving Multi-Modal Planning and Shared Use of Infrastructure”.
See More
Project Details
| Project Type: | Dissertation |
|---|---|
| Project Status: | Completed |
| End Date: | December 05, 2018 |
| UTC Funding: | $15,000 |
Downloadable Products
- Combined Model Approach to the Problem of Ranking (FINAL_REPORT)