🌦️ Deep Learning Project: RNN for Seattle Weather Prediction

📑 Project Description

This project utilizes a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) to predict future maximum temperatures based on historical weather data from Seattle. The project includes data preprocessing, model training, hyperparameter optimization, and visualization of predictions.

📂 Project Structure

.
├── Dataset.py          # Data loading and preprocessing
├── Model_RNN.py        # RNN model construction
├── Optimize.py         # Hyperparameter optimization with Keras Tuner
├── Train.py            # Model training script
├── Test.py             # Model evaluation
├── Visualisation.py    # Visualization and image saving
├── Logger.py           # Custom logger with detailed and humorous logs
├── Main.py             # Main script to run the project
├── requirements.txt    # Python dependencies
├── plots/              # Folder to save generated prediction plots
└── app.log             # Log file capturing all project logs

🚀 Installation

# Clone the repository
git clone https://github.com/your-repo/rnn-weather-prediction.git
cd rnn-weather-prediction

# Create a virtual environment
python -m venv env
source env/bin/activate  # On Windows: env\Scripts\activate

# Install dependencies
pip install -r requirements.txt

🏃 Usage

python Main.py

• Logs will be recorded in app.log.
• Prediction plots will be saved automatically in the plots directory.

📈 Output Examples

• rnn_weather_prediction.png : Displays the predicted vs actual maximum temperatures.
• app.log : Contains detailed logs including training progress, hyperparameter tuning, and potential errors.

⚙️ Dependencies

pandas==1.5.3
numpy==1.24.3
scikit-learn==1.2.2
tensorflow==2.12.0
matplotlib==3.7.1
keras-tuner==1.3.5

🧠 Optimizing the RNN Model

The project uses Keras Tuner to find the best hyperparameters for the RNN model, including:

• LSTM units: Between 32 and 128
• Learning rate: Choices of 0.01, 0.001, and 0.0001

🔥 Example Logs

📝 INFO: Starting the application for RNN weather prediction 🌦️.
🔍 DEBUG: Successfully loaded dataset! Size: (1461, 6).
🚨 ERROR: File not found: /mnt/data/seattle-weather.csv
🎯 INFO: Model training completed! Ready to predict the future.

🙋 Support

If you encounter any issues, please open an issue or contact me directly.

---

💡 Tip: Always activate your virtual environment before running the project!