Summary of Robust Time Series Forecasting with Non-heavy-tailed Gaussian Loss-weighted Sampler, by Jiang You et al.
Robust Time Series Forecasting with Non-Heavy-Tailed Gaussian Loss-Weighted Sampler
by Jiang You, Arben Cela, René Natowicz, Jacob Ouanounou, Patrick Siarry
First submitted to arxiv on: 19 Jun 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 This paper introduces a novel approach to forecasting multivariate time series, addressing challenges posed by extreme or redundant samples. Recent resampling methods aim to increase training efficiency by reweighting samples based on their running losses, but these methods do not solve the problems caused by heavy-tailed distribution losses. The proposed Gaussian loss-weighted sampler multiplies the running losses with a Gaussian distribution weight, reducing the probability of selecting samples with very low or high losses and favoring those close to average losses. This approach creates a weighted loss distribution that is not heavy-tailed theoretically, offering several advantages over existing methods. Applications on real-world time series forecasting datasets demonstrate improvements in prediction quality using mean square error measurements. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us make better predictions about what will happen in the future by dealing with tricky data problems. Sometimes, data can be weird and have really high or low values that mess up our predictions. The authors created a new way to look at this data, called a Gaussian loss-weighted sampler. This tool makes it more likely for us to use the good data and less likely to use the bad data. As a result, we get better predictions! |
Keywords
* Artificial intelligence * Probability * Time series
