Using information theory to detect model structure with application in vehicular traffic systems

Lane, Daniel and Roy, Subhradeep Using information theory to detect model structure with application in vehicular traffic systems. Using information theory to detect model structure with application in vehicular traffic systems. pp. 367-372.

Text 3. jurnal 3.pdf - Published Version Download (718kB)

Abstract

As with other complex dynamic systems, modeling traffic systems requires an accurate understanding of how the individual components are coupled, which can be challenging without prior knowledge. Due to a lack of understanding of the nature of interactions between traffic entities, existing models rely on assumptions. The present study evaluates two information-theoretic measures and their ability to quantify and determine the nature of interactions using synthetic data generated from two structurally different traffic simulation models. These measures uncover relationships that describe these interactions without knowing the underlying model, which suggests these measures can be useful in data-driven model discovery

Actions (login required)

View Item