Title:
AUTOMATIC TRAINING-BASED TIME SERIES DATA PREDICTION AND CONTROL METHOD AND APPARATUS
Document Type and Number:
WIPO Patent Application WO/2022/145981
Kind Code:
A1
Abstract:
Disclosed are an automatic training-based time series data prediction and control method and apparatus. The automatic training-based time series data prediction and control method according to an embodiment comprises the steps of: training each of a plurality of time series data prediction models according to a condition for each corresponding model; determining one or more optimal models satisfying a predetermined condition, from among the trained time series data prediction models; and combining the one or more optimal models to generate a final model, wherein the plurality of time series data prediction models include at least one of a statistical-based prediction model and a deep learning-based prediction model.
Inventors:
KIM SONG HWAN (KR)
Application Number:
PCT/KR2021/020067
Publication Date:
July 07, 2022
Filing Date:
December 28, 2021
Export Citation:
Assignee:
INEEJI (KR)
International Classes:
G06Q10/04; G05B13/04; G06N3/04; G06N3/08
Foreign References:
| KR20200017604A | 2020-02-19 | |||
| KR102037279B1 | 2019-11-15 | |||
| KR20200014510A | 2020-02-11 | |||
| KR20170078256A | 2017-07-07 | |||
| KR102041545B1 | 2019-11-06 |
Other References:
See also references of EP 4075353A4
Attorney, Agent or Firm:
MUHANN PATENT & LAW FIRM (KR)
Download PDF: