| 摘要 |
A time-series model learning unit (11) reads in the time-series data sequentially, learns the parameters of a time-series model and stores the learned parameters in a complexity and parameter storage device (14). A loss function calculating unit(13) reads in sequentially from the storage device with regard to each item of the input time-series data one at a time and calculates the values of a loss function. A complexity calculating unit (12) sums the loss values to sequentially calculate complexity as fitting error that results when a time-series model is fit to the input time-series data. Complexity is stored in the storage device. A change-point searching unit (15) reads in complexity regarding time-series data before and after change-point candidates from the storage device with respect to all change-point candidates, compares the sum of the complexities with complexity regarding all time-series data in a case where a change point is assumed to be absent, calculates the change-point score based upon the difference between the compared values, detects a change point in accordance with the magnitude of the change-point score and outputs the change point.
|
