Prediction device, prediction method, and recording medium
AssigneeCanon Kabushiki Kaisha
InventorsHiromi Shiozaki, Atsushi Nogami
Life of this patent
Estimating the term…
Abstract
To precisely predict future data even when the number of pieces of time-series data is small, in predicting the future data, using the time-series data. When the future data is predicted using the time-series data, whether present time data is used is determined based on prediction variation or a data transition, and then the prediction of the future data is performed.
Claims15
A prediction device comprising:
- one or more processors;
- and
- a memory having instructions stored thereon which, when executed by the one or more processors, cause the device to:
- acquire present time data;
- generate time-series data from the data acquired at a plurality of times;
- calculate a prediction result in a present time from the present time data and the past time data;
- determine whether the generated time-series data satisfies a predetermined condition based on a first difference between a prediction result calculated in the present time and a prediction result in a past time calculated in a past, and a second difference between the prediction results in past times calculated in the past;
- and
- predict future data, based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and to predict the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
A prediction device comprising:
- one or more processors;
- and
- a memory having instructions stored thereon which, when executed by the one or more processors, cause the device to:
- acquire present time data;
- generate time-series data from the data acquired at a plurality of times;
- calculate a prediction result in a present time from the present die data and the past time data;
- detect a transition of data similar to a transition of a part of the time-series data from past time-series data;
- and
- determine whether the generated time-series data satisfies a predetermined condition based on the detected similar transition of data;
- and
- predict future data, a based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and to predict the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
A prediction method comprising:
- acquiring present time data;
- generating time-series data from the data acquired at a plurality of times;
- calculate a prediction result in a present time from the present time data and the past time data;
- determining whether the generated time-series data satisfies a predetermined condition based on a first difference between a prediction result calculated in the present time and a prediction result in a past time calculated in a past, and a second difference between the prediction results in past times calculated in the past;
- and
- predicting future data, based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and predicting the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
A non-transitory computer-readable recording medium that stores a program for causing a computer to function as a prediction device configured to:
- acquire present time data;
- generate time-series data from the data acquired at a plurality of times;
- calculate a prediction result in a present time from the present time data and the past time data;
- determine whether the generated time-series data satisfies a predetermined condition based on a first difference between a prediction result calculated in the present time and a prediction result in a past time calculated in a past, and a second difference between the prediction results in past times calculated in the past;
- and
- predict future data, based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and to predict the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
A prediction method comprising:
- acquiring present time data;
- generating time-series data from the data acquired at a plurality of times;
- calculating a prediction result in a present time from the present time data and the past time data;
- detecting a transition of data similar to a transition of a part of the time-series data from past time-series data;
- and
- determining whether the generated time-series data satisfies a predetermined condition based on the detected similar transition of data;
- and
- predicting future data, based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and to predict the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
A non-transitory computer-readable recording medium that stores a program for causing a computer to function as a prediction device configured to:
- acquire present time data;
- generate time-series data from the data acquired at a plurality of times;
- calculate a prediction result in a present time from the present time data and the past time data;
- detect a transition of data similar to a transition of a part of the time-series data from past time-series data;
- and
- determine whether the generated time-series data satisfies a predetermined condition based on the detected similar transition of data;
- and
- predict future data, based on past time data without using the present time data, when the predetermined condition is determined to be satisfied, and to predict the future data, based on the present time data and the past time data, when the predetermined condition is determined not to be satisfied,
- wherein past time data is previously acquired present time data which has been acquired prior to the present time data.
Dependencies are inferred from the claim text.