Method, apparatus and program for combining various spatial resolution and time resolution forecast

摘要

Problem to be solved: to provide a method, apparatus and program for combining various spatial resolution and time resolution forecasts, which can be seamless and can suppress the reduction in accuracy.The method, apparatus, and the program acquire multiple learning forecasts at a plurality of time steps at at least one point of a plurality of points.A plurality of learning states are obtained at a plurality of time steps at one point.The machine learning algorithm is trained so that the integrated forecast is output based on multiple learning forecasts and multiple learning conditions.The trained machine learning algorithm is implemented to output an integrated forecast based on multiple forecasts.Here, multiple learning forecasts and multiple forecasts are different from each other at least one of the spatial resolution or time resolution.Further, a plurality of learning forecasting and multiple forecasts include a first forecast effective in a first time step and a second forecast effective in a second time step.Diagram