Time series forecasting

Patent metadata

Field	Value
Publication number	US-2024362458-A1
Application number	US-202318309268-A
Country	US
Kind code	A1
Filing date	Apr 28, 2023
Priority date	Apr 28, 2023
Publication date	Oct 31, 2024
Grant date	—

Abstract

A method, system, and computer program product that is configured to: receive an input time series from an external device in a first system, divide the input time series to a set of univariate time subseries, transform the set of univariate time subseries into a univariate prediction result series using a transformer model, concatenate the univariate prediction result series to a multivariate predictive result, and output the multivariate predictive result for providing time series forecasting to a second system.

First claim

What is claimed is:

1 . A method, comprising:
receiving, by a processor set, an input time series from an external device in a first system; dividing, by the processor set, the input time series to a set of univariate time subseries; transforming, by the processor set, the set of univariate time subseries into a univariate prediction result series using a transformer model; concatenating, by the processor set, the univariate prediction result series to a multivariate predictive result; and outputting, by the processor set, the multivariate predictive result for providing time series forecasting to a second system.

2 . The method of claim 1 , wherein the external device comprises a smart sensor, the first system comprises a manufacturing system, and the second system comprises a planning system in communication with the first system.

3 . The method of claim 1 , wherein the input time series comprises a multivariate time series.

4 . The method of claim 3 , wherein the multivariate time series comprises a multi-channel signal.

5 . The method of claim 1 , wherein the transforming the set of univariate time subseries into the univariate prediction result series comprises normalizing and segmenting the univariate time subseries into patches.

6 . The method of claim 5 , wherein the patches are local and semantic information in aggregated time steps.

7 . The method of claim 5 , wherein the transforming the set of univariate time subseries into the univariate prediction result series further comprises transforming the patches into a representation.

8 . The method of claim 7 , wherein the transforming the set of univariate time series into the univariate prediction result series further comprises utilizing a flatten layer with a linear head on the representation to obtain the univariate prediction result series.

9 . The method of claim 1 , wherein the univariate time subseries comprises a plurality of channel independent signals.

10 . The method of claim 9 , wherein each of the channel independent signals have a same model weight as a weight of remaining channel independent signals.

11 . The method of claim 1 , wherein the transformer model comprises a supervised model.

12 . A computer program product comprising one or more computer readable storage media having program instructions collectively stored on the one or more computer readable storage media, the program instructions executable to:
receive an input time series from an external device in a first system; divide the input time series to a set of univariate time subseries; pre-train a transformer model using historically reconstructed masked patches; transform the set of univariate time subseries into a univariate prediction result series using the pre-trained transformer model; concatenate the univariate prediction result series to a multivariate predictive result; and output the multivariate predictive result for providing time series forecasting to a second system.

13 . The computer program product of claim 12 , wherein the external device comprises a smart sensor, the first system comprises a manufacturing system, and the second system comprises a planning system in communication with the manufacturing system.

14 . The computer program product of claim 12 , wherein the input time series comprises a multivariate time series.

15 . The computer program product of claim 14 , wherein the multivariate time series comprises a multi-channel signal.

16 . The computer program product of claim 15 , wherein the transforming the set of univariate time subseries into the univariate prediction result series comprises normalizing, segmenting, and masking the univariate time subseries into masked patches and non-masked patches.

17 . The computer program product of claim 16 , wherein the transforming the set of univariate time subseries into the univariate prediction result series further comprises utilizing a linear layer on the non-masked patches to obtain the univariate prediction result series.

18 . The computer program product of claim 17 , wherein the transforming the set of univariate time subseries into the univariate prediction results series further comprises reconstructing the masked patches.

19 . The computer program product of claim 12 , wherein the univariate time subseries comprises a plurality of channel independent signals.

20 . The computer program product of claim 19 , wherein each of the channel independent signals have a same model weight as a weight of remaining channel independent signals.

21 . The computer program product of claim 12 , wherein the transformer model comprises a self-supervised model.

22 . A system comprising:
a processor set, one or more computer readable storage media, and program instructions collectively stored on the one or more computer readable storage media, the program instructions executable to: receive an input time series from an external device in a first system; divide the input time series to a set of univariate time subseries; pre-train a transformer model using historically reconstructed masked patches; transform the set of univariate time subseries into a univariate prediction result series using the pre-trained transformer model; concatenate the univariate prediction result series to a multivariate predictive result; and output the multivariate predictive result for providing time series forecasting to a second system.

23 . The system of claim 22 , wherein the transforming the set of univariate time subseries into the univariate prediction result series comprises normalizing, segmenting, and masking the univariate time subseries into masked patches and non-masked patches.

24 . A method, comprising:
receiving, by a processor set, a univariate time series; dividing, by the processor set, the univariate time series into patches; transforming, by the processor set, the patches into a representation using a transformer model; obtaining, by the processor set, a univariate prediction result series by using a flatten layer with a linear head on the representation; and outputting, by the processor set, the univariate prediction result series.

25 . A computer program product comprising one or more computer readable storage media having program instructions collectively stored on the one or more computer readable storage media, the program instructions executable to:
receive a univariate time series; divide the univariate time series into a non-overlapped set of patches; mask a subset of the non-overlapped set of patches to a masked patch series; pre-train a transformer model using historically reconstructed masked patches: transform the non-overlapped set of patches to a univariate prediction result series using the pre-trained transformer model; and output the univariate prediction result series.

Assignees

• IBM

Inventors

• Nguyen Nam H
• Nie Yuqi
• Reddy Chandrasekhara K
• Patel Dhavalkumar C
• Bhamidipaty Anuradha
• Kalagnanam Jayant R
• Sinthong Phanwadee

Classifications

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• G06N3/0455 Primary

  Auto-encoder networks; Encoder-decoder networks · CPC title

• G06N3/0895 Primary

  Weakly supervised learning, e.g. semi-supervised or self-supervised learning · CPC title