Predictive model generation system and predictive system

摘要

PROBLEM TO BE SOLVED: To accurately predict statistical values of people's social life by using information on people's movement. A predictive model generation system 1 is a system that generates a predictive model that predicts activity statistics associated with a place, and is based on inter-place relationship information indicating the number of people moving between places. By reflecting the actual statistics of other places in the actual statistics, which are the actual values of the past activity statistics before the reference time at the place of, the feature data showing the characteristics of the activity statistics of one place is generated. It was obtained by inputting the feature data into the prediction model using the first feature data generation unit 14, the learning data consisting of the feature data of each place and the correct answer data consisting of the actual statistical values at the reference time of each place. It includes a prediction model generation unit 15 that generates a prediction model by updating the parameters of the prediction model based on the error between the model output value and the correct answer data. [Selection diagram] Fig. 1