Electricity Price Forecast
EPF (Electricity Price Forecasting) is a full-stack machine learning application that predicts Spanish day-ahead electricity market prices up to 7 days (168 hours) into the future. It collects real-time data from the Spanish electricity system operator (REE/ESIOS), weather services (AEMET), and commodity markets (TTF gas, ETS carbon), transforms them into 50+ engineered features, and feeds them into gradient boosting models to produce hourly price forecasts.
Electricity prices in liberalised markets, such as Spain's OMIE (Operador del Mercado Ibérico de Energía), are highly volatile. They are driven by a complex interplay of demand patterns, renewable generation (wind and solar are inherently variable), fuel costs (natural gas, carbon permits), interconnection flows, and calendar effects (holidays, weekends, vacation periods). Accurate short-term price forecasting benefits energy traders, renewable producers, grid operators, and end consumers.
