Inter-Community Commuting Patterns Web Visualization
Driver Identification using Trajectory Data
In this paper, we explore the possibility of constructing driver
identification model using the route and driving time
information which uses the trajectory data as the only input. We
use dataset from Safety Pilot Model Deployment (SPMD) and
matched the original trajectory to the existing road network
using Hidden Markov Model (HMM). To form the training data,a
reformatted matrix is created to reflect the distance traveled
in each network link and the travel time in a day. Both Nearest
Subspace (NS) and Random Forest (RF) are trained in the training
stage. [PDF]
Location Quotient Web App
Visualize the dominant industries in each state and county using
Location quotient