You can read more about the project on Medium.
A data science project aimed at understanding traffic congestions in terms of duration and its anomalies primarily in WA state. Pipelines are built to be but scalable to all of US.
The project is twofold: a regression model to predict the duration of congestion if it happens, leading into a anomaly detector to see if early intervention is necessary.
The app developed for this project is not deployed due to cost concerns, you can find the app in "TrafficApp" folder.
Primary data source: “Short and Long-term Pattern Discovery Over Large-Scale Geo-Spatiotemporal Data.”
├── LICENSE
├── README.md
├── data
│ ├── processed <- Processed traffic, weather, population data.
│ ├── interim <- Intermediate data that has been transformed.
│ └── raw <- Raw traffic and weather event data.
│
├── models <- Contains trained regression model and classifier.
│
├── notebooks <- The code written for this project, ordered.
│
├── presentation <- Presentation and video made for this project.
│
├── TrafficApp <- Contains the live bus tracker developed for this project.
│
└── requirements.txt <- The requirements file for reproducing the analysis environment.
Useful resources to build the app:
Thank you.