VGU RESEARCH REPOSITORY
Please use this identifier to cite or link to this item:
https://epub.vgu.edu.vn/handle/dlibvgu/1716| Title: | Self-supervised learning for time series forecasting | Authors: | Nguyen Khoa | Keywords: | Time series forecasting;Self-supervised learning;Zero-shot learning;Transfer learning | Issue Date: | 2023 | Abstract: | In the realm of time series forecasting, the challenge often revolves around limited data availability. This thesis, titled "Self-supervised Learning For Time Series Forecasting," delves into this issue by harnessing the power of self-supervised learning techniques. The study introduces the Enhanced Rolling Windows technique, which accelerates data splitting for larger datasets. A novel linear model, drawing from RevIN and Time2Vec, is also proposed. Additionally, the research explores advanced concepts like Zero-Shot Learning and Transfer Learning, highlighting their significance in forecasting. These methodologies are particularly applied to forecast salinity datasets. In essence, the thesis presents a comprehensive strategy for time series forecasting, spotlighting the importance of self-supervised techniques in navigating the challenges of limited data. |
URI(1): | https://epub.vgu.edu.vn/handle/dlibvgu/1716 | Rights: | Attribution-NonCommercial 4.0 International |
| Appears in Collections: | Computer Science (CS) |
Files in This Item:
| File | Description | Size | Format | Existing users please Login |
|---|---|---|---|---|
| Self-supervised learning for time series forecasting.pdf | 6.48 MB | Adobe PDF |
Page view(s)
149
checked on Oct 16, 2025
Download(s)
47
checked on Oct 16, 2025
Google ScholarTM
Check
This item is licensed under a Creative Commons License