This repository contains the starting code for the SMP forecasting project. The goal is to develop a comprehensive price forecasting system for Turkey's electricity market using various machine learning techniques.

• data/: Raw and processed data files
• notebooks/: Jupyter notebooks for exploration and analysis .
• src/: Source code for the project
• tests/: Unit tests

1. Clone the repository
2. Install dependencies: pip install -r requirements.txt
3. Follow the notebooks in order:
   • Data collection
   • Data exploration
   • Model evaluation

See the project documentation for detailed requirements and grading criteria.

• Statistical Models (ARIMA, SARIMA, PROPHET)
• Machine Learning Models (SVR, Random Forest, XGBoost)
• Deep Learning Models (LSTM, Transformer)

Implement the same models with additional features:

• Weather data
• Currency exchange rates
• Economic indicators
• Production sources

Models will be evaluated using:

• MAPE (Mean Absolute Percentage Error)
• RMSE (Root Mean Square Error)
• MAE (Mean Absolute Error)