Taming the Wind: Fast ML‑Calibrated Seasonal Models for Wind‑Power Derivatives

This master’s thesis develops a practical toolkit for pricing and risk management of wind-power derivatives by enhancing a seasonal, mean-reverting stochastic model with an explicit market-price-of-risk component to ensure both analytical tractability and consistency with observed market behavior. To fit the model to real-world data, two advanced machine-learning calibration techniques are implemented and compared in terms of accuracy, computational speed, and robustness.