🚦 Traffic Flow Prediction Using Neural Networks 🧠🚗

Lab 5 Report by: Donfack Tsopfack Yves Dylane
Instructor: Mbachan Fabrice
📍 City Focus: Yaoundé, Cameroon

---

🌟 Introduction

Traffic congestion is one of the most pressing issues in modern cities 🌍, and Yaoundé is no exception. With increasing vehicle density 🚘 and complex traffic patterns, managing traffic efficiently is a high priority. This project leverages the power of Neural Networks 🧠 to predict traffic flow and congestion patterns, showcasing how AI can drive smart city solutions 🚀.

In this lab, we explored:

• 🌦️ How weather and time-related factors impact traffic.
• 📊 Techniques for handling sequential data.
• 🛠️ Building and training neural networks for prediction.

This project isn't just about data—it's about creating smarter cities for better urban mobility 🏙️.

---

🎯 Objective

• 📈 Predict traffic congestion levels using historical data, weather conditions, and time-related features.
• 🧠 Build and train neural networks for forecasting.
• 🔍 Explore spatial and temporal traffic trends for smarter city solutions.

---

🧑‍💻 Skills Developed

• 📅 Handling time-series (sequential) data.
• 🛠️ Building predictive models using TensorFlow/Keras.
• 🔄 Identifying patterns in traffic influenced by environmental factors.

---

⚙️ Tools & Technologies

• Python Libraries:
  • 📊 NumPy & Pandas: Data manipulation experts!
  • 🧠 Keras & TensorFlow: The brains of our neural network.
  • 🎨 Matplotlib & Seaborn: Data visualization champions.
  • 🛠️ Scikit-learn: Helper for preprocessing and performance metrics.
• Dataset: Historical traffic, weather 🌦️, and time ⏰ data (or simulated if unavailable).

---

🚀 Project Workflow

🔢 1. Data Preprocessing

• ✅ Loaded data and cleaned missing entries.
• 📅 Extracted time-based features: hour, day, weekend labels.
• ⚖️ Normalized weather data (e.g., temperature, humidity).

🎨 2. Feature Engineering

• 🕒 Added lag features (e.g., traffic from previous hours).
• 🛠️ Prepared data for model input with MinMaxScaler.

✂️ 3. Data Splitting

• 📂 Split data into 80% training, 20% testing, with optional validation set.
• ⚖️ Ensured all data was consistently scaled for better performance.

🏗️ 4. Model Building

• Designed a Sequential Neural Network with:
  • 🔄 LSTM Layers for sequence learning.
  • 🚪 Dropout Layers to prevent overfitting.
  • 📊 Dense Layers for refined predictions.
• Loss Function: Mean Squared Error (MSE).
• Optimizer: Adam 🛠️.

🧪 5. Model Training & Evaluation

• 🏋️‍♂️ Trained the model for 50 epochs with early stopping and learning rate reduction.
• 📊 Evaluated with metrics:
  • 🛠️ MSE, MAE for error analysis.
  • 🔍 MAPE for percentage accuracy.

---

📊 Results

Metrics Achieved:

• 🟢 Validation Loss (MSE): 0.0108.
• 🟢 Validation MAE: 0.0809.
• 🟢 Validation MAPE: 24.4%.

💡 Insights:

• Time features like rush hour and weekends heavily influence traffic flow.
• Environmental conditions like rain 🌧️ and temperature 🌡️ also play a significant role.

---

🔮 Future Scope

• 🛣️ Expand the model to include real-time traffic data.
• 🌐 Integrate with APIs for live weather updates.
• 🧠 Enhance prediction accuracy with more advanced architectures (e.g., Transformer models).

---

👨‍💻 How to Run

1. 📥 Clone the repo:

   git clone https://github.com/yvesdylane/Traffic_Flow_Prediction_Using_Neural_Networks

2. 📂 Navigate to the directory:

   cd Traffic_Flow_Prediction_Using_Neural_Networks

3. ⚙️ Install dependencies:

   pip install -r requirements.txt  

4. 🏃‍♂️ Run the Complete Main file ✨.

🤝 Contributions

Got ideas? Found a bug 🐛? Submit an issue or create a pull request!

📄 License

This project is licensed under the MIT License.

🎉 Thank you for visiting! Don't forget to ⭐ this repo if you find it helpful!