This repository contains the code files produced during an extensive research in the field of mobility data analytics.
- Objective: We are trying to predict the forthcoming traffic flow volume for specific routes within San Francisco, California, leveraging insights gleaned from historical data.
- Methodology: We perform time series forecasting using machine and deep learning models. We use the Strict Path Queries algorithm to measure traffic flow in each path accurately.
- Data: We use traffic flow data of Yellow Taxis that are moving within the city or San Francisco, California. We use this original dataset to generate the final time series data. This entire process is described in the Jupyter notebook files inside the Python_Code folder. Notebooks have increment numbers that define the order of their execution.
- Number of Paths: A total of 100 paths (or pathways) are used for conducting forecasts.
To run the code in this repository, ensure you have the latest version of Python installed. The required libraries are listed in the Necessary_Libraries.txt file. You can install them using pip or conda commands.
My name is Efstratios Karkanis, and I have finished my studies in computer science at the University of Piraeus. For any inquiries or to establish contact, please feel free to reach out to me at stratoskarkanis2@gmail.com.
Feel free to explore the code and the insights gained from this project. Contributions and feedback are always welcome!