Summary of The Impact Of Data Set Similarity and Diversity on Transfer Learning Success in Time Series Forecasting, by Claudia Ehrig et al.
The impact of data set similarity and diversity on transfer learning success in time series forecasting
by Claudia Ehrig, Benedikt Sonnleitner, Ursula Neumann, Catherine Cleophas, Germain Forestier
First submitted to arxiv on: 9 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary In this research paper, the authors investigate how pre-trained models can be used for time series forecasting on target data sets. They focus on understanding which characteristics of the source and target data are most important for successful transfer learning. The study uses five public source datasets to predict five different target datasets, including real-world wholesale data. The results show that source-target similarity reduces bias in forecasting, while source diversity improves accuracy and uncertainty estimation. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper explores how pre-trained models can be used for time series forecasting on target data sets by understanding the importance of source and target data characteristics. They use five public source datasets to predict five different target datasets, including real-world wholesale data. The results show that source-target similarity reduces bias in forecasting, while source diversity improves accuracy and uncertainty estimation. |
Keywords
* Artificial intelligence * Time series * Transfer learning
