Search for: All records

In projects centered around rare event case data, the challenge of data comprehension is greatly increased because of insufficient data for deriving insight and analysis. This is particularly the case with traffic crash occurrence, where positive events (crashes) are rare and, in most cases, no data set exists for negative events (non-crashes). One method to increase available data is negative sampling, which is the process of creating a negative event based on the absence of a positive event. In this work, four negative sampling techniques are presented with varying ratios of negative to positive data. These types of techniques are based on spatial data, temporal data, and a mixture of the two, with the data ratios acting as class balancing tools. The best performing model found was with a negative sampling technique that shifted temporal information and had an even 50/50 data split, with an F-1 score, a formulaic combination of precision and recall, of 93.68. These results are promising for Inteligent Transportation Systems (ITS) applications to inform of potential crash locations in an entire area for proactive measures to be put in place. 

Autonomous vehicle (AV) fleet management is one of the major aspects of AV development that needs to be standardized before AV deployment. There has been no consensus on whether AV deployment in general will be beneficial or detrimental in terms of road congestion. There are similarities between packet transmission in computer networks and AV transportation in road networks. In this work, the authors argue that congestion avoidance algorithms used in computer networks can be applied for AV fleet management. Authors modify and evaluate a novel adaptation of additive increase and multiplicative decrease (AMID) congestion avoidance algorithm. The authors propose assigning different priorities to transportation tasks in order to facilitate sharing the limited resources in such as usage of the road network. This will be modeled and assessed using a queueing model based on AVs arrival distribution. This will result in a load balancing paradigm that can be used to share and manage limited resources. Then, by using numerical study authors merge congestion avoidance and load balancing to analyze the authors' scheme in term of road network throughput (number of cars in network for a given time) for AV fleet management. Their evaluation demonstrates the improvement in terms of road network throughput.