A data processing method and apparatus
By converting time-series information to the frequency domain, selecting a portion of the frequency bands for feature extraction, and then converting it back to the time domain, combined with multilayer perceptron (MLP) processing, the problem of insufficient processing capability of temporal convolutional networks in existing technologies is solved, achieving more efficient sequence information processing.
CN117391138BActive Publication Date: 2026-05-29HUAWEI TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUAWEI TECH CO LTD
- Filing Date
- 2023-09-28
- Publication Date
- 2026-05-29
Smart Images
Figure CN117391138B_ABST
Abstract
A data processing method, the method comprising: obtaining time series information; processing the time series information through an encoder to obtain an encoding result; wherein the encoder comprises a target network layer; the target network layer is configured to convert features of the time series information to a frequency domain to obtain frequency spectrum information; the frequency spectrum information comprises a plurality of frequency bands and frequency domain features corresponding to the frequency bands; according to the frequency domain features, a part of the frequency bands is selected from the plurality of frequency bands, and feature extraction is performed on the frequency domain features of the part of the frequency bands to obtain a processing result, and the processing result is converted to a time domain; and according to the encoding result, a task network is used to obtain a task processing result. The application improves the processing performance of the encoder on the sequence information by modeling the frequency domain features.
Need to check novelty before this filing date? Find Prior Art