| 摘要 |
PROBLEM TO BE SOLVED: To efficiently generate the teacher data used in the learning of a machine learning system by classifying the data extracted from time-series data on the basis of specific standard.SOLUTION: A data processing device according to the present invention comprises: a data extraction unit for extracting, from time-series data, a teacher data candidate that is data at specific timing; a teacher data generation unit for generating teacher data on the basis of a label, by which the teacher data candidate can be classified, and the teacher data candidate; and a teacher data complement unit for further extracting, on the basis of the extent of a change between a specific teacher data candidate and other teacher data candidates preceding the specific teacher data candidate in the time series, a teacher data candidate from the time-series data present between the specific teacher data candidate and the other teacher data candidates, the teacher data generation unit assigning a label to the extracted teacher data candidate and adding this data to the teacher data.SELECTED DRAWING: Figure 13 |
