Natural gas demand forecasting for business customers
Prognozowanie zużycia gazu naturalnego przez klientów biznesowych
Piotr Pawłowski
Streszczenie
The paper presents results for business clients portfolio natural gas demand forecasting. Historical data set was
analysed to find dependences between gas demand and any other variables such as weather variables and some social
behaves of gas customers. Advanced statistical methods have been applied to achieve high model accuracy. Article presents
two approaches to forecasting model design. First one forecasts whole portfolio gas demand and the other one forecasts each
individual customer gas demand and then summarizes it into the whole portfolio gas demand. The article clearly disclose
how to create a forecasting model and which approach is more accurate.
analysed to find dependences between gas demand and any other variables such as weather variables and some social
behaves of gas customers. Advanced statistical methods have been applied to achieve high model accuracy. Article presents
two approaches to forecasting model design. First one forecasts whole portfolio gas demand and the other one forecasts each
individual customer gas demand and then summarizes it into the whole portfolio gas demand. The article clearly disclose
how to create a forecasting model and which approach is more accurate.