AI Project: Austin Traffic

Download the following files to the data folder.

Volume data

Link to dataset source: https://data.austintexas.gov/Transportation-and-Mobility/Traffic-Studies-Vehicle-Volume-Reports-BETA-/jasf-x4rx Download link: https://data.austintexas.gov/api/views/jasf-x4rx/rows.csv?accessType=DOWNLOAD

Location data

Link to dataset source: https://data.austintexas.gov/Transportation-and-Mobility/Traffic-Studies-Locations-BETA-/jqhg-imb3 Download link: https://data.austintexas.gov/api/views/jqhg-imb3/rows.csv?accessType=DOWNLOAD

Running the notebook

Make sure you have Docker installed and running. Then change the path to this repository in start_docker.sh and run the script.

Important Files

Loading, cleaning and prepping the data for the bayesion network can be found in dockless_data.py. The code for the bayesian network and learning the conditional probablities as well as sampling are in dockless_model.py. The simulator that computes the paths and stores the frequencies in a networkx graph is found in simluator.py.

The notebooks DocklessScooterBayes.ipynb and DocklessScooterBayes[Prototype].ipynb detail how we experiemented and created the bayes net.