METHOD AND SYSTEM FOR WEATHER FORECAST CREATION Russian patent published in 2017 - IPC G01W1/10 G06Q90/00 

FIELD: information technology.

SUBSTANCE: method is run on a server that includes a processor and a machine learning module. The method includes: obtaining by the machine learning module the current weather measurement parameter representing the weather parameter at the measurement timepoint; obtaining by the machine learning module the first average value of the historical weather parameter for the measurement timepoint; creation of the normalised value of the weather measurement parameter by the machine learning module, on the basis of the difference between the current weather measurement parameter and the first average historical weather parameter for the measurement timepoint; training the machine learning module to create a normalised value of the weather prediction parameter, at least partially, on the basis of the normalised value of the weather measurement parameter, the normalised value of the weather prediction parameter is linked with the prediction timepoint after the measurement timepoint.

EFFECT: increased accuracy of forecasting weather parameters by reducing the prediction performance ratio.

45 cl, 7 dwg