As a Machine Learning Project, CE Department of Amirkabir University of Technology, Fall 2021
This Python code simulates the daily passenger flow within a metro system represented as a directed graph. The simulation generates a dataset capturing the in-flow, out-flow, and line traffic at each station over a specified period. Additionally, the code demonstrates the use of machine learning models, specifically Linear Regression and AdaBoostRegressor, to predict passenger in-flow, out-flow, and line traffic in the metro system.
numpypandasnetworkxmatplotlibscikit-learn
Make sure to install the required dependencies before running the code.
pip install numpy pandas networkx matplotlib scikit-learn- The metro system is represented as a directed graph using the NetworkX library.
- Station positions are set, and edges between stations represent metro lines with distinct colors.
- Station rates are defined based on in-degree, out-degree, and the existence of multiple lines.
- The simulation generates a dataset capturing daily passenger flow, considering station connectivity and line traffic.
- The code visualizes input, output, and line traffic for selected stations on the first day.
- One-hot encoding is applied to station names for machine learning model input.
- Linear Regression models are trained to predict passenger in-flow and out-flow.
- An AdaBoostRegressor model is trained to predict line traffic.
- Examples of predictions for specific times and stations are provided.
- The average in-flow and out-flow rates for each station are estimated.
- Install the required dependencies.
- Run the code to simulate metro system traffic and train machine learning models.
- Analyze the visualizations and model predictions.
Feel free to customize the code for your specific metro system or use case.