Title | ||
---|---|---|
A large scale examination of vehicle recorder data to understand relationship between drivers' behaviors and their past driving histories |
Abstract | ||
---|---|---|
We propose an analysis method of driving behaviors based on large-scale and long-term vehicle recorder data to support fleet driver management by classifying drivers by their skill, safety, physical/mental fatigue, aggressiveness, and so on. Previous studies rely on precise data with small number of drivers, which are difficult to extrapolate to general drivers. In this study, we examine ability of a dataset that is sparse but large-scale (over 100 fleet drivers) and long-term (10 months' worth). We focus on classifying drivers recently involved in accidents, and examine correlation with driving behaviors. We propose two models for the classification; entropy-like model and KL divergence model that aim to emphasize the behavioral difference from average drivers. From experiments, we will show some informative findings on behaviors that might cause accidents. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1109/BigData.2015.7364108 | Big Data |
Field | DocType | Citations |
Mental fatigue,Computer science,Correlation,Artificial intelligence,Machine learning,Kullback–Leibler divergence | Conference | 2 |
PageRank | References | Authors |
0.43 | 7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Daisaku Yokoyama | 1 | 71 | 9.77 |
Masashi Toyoda | 2 | 388 | 49.87 |