Summary of Frequency Adaptive Normalization For Non-stationary Time Series Forecasting, by Weiwei Ye et al.
Frequency Adaptive Normalization For Non-stationary Time Series Forecasting
by Weiwei Ye, Songgaojun Deng, Qiaosha Zou, Ning Gui
First submitted to arxiv on: 30 Sep 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 The proposed frequency adaptive normalization (FAN) method addresses non-stationarity in time series forecasting by extending instance normalization to handle both dynamic trends and seasonal patterns. The approach employs Fourier transform to identify predominant frequency components, which are then modeled as a prediction task with a simple MLP model. FAN is applied to arbitrary predictive backbones, demonstrating significant performance advancements on eight benchmark datasets, with average improvements ranging from 7.76% to 37.90% in MSE. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Time series forecasting is important for predicting what will happen next based on past data. However, this type of data can be tricky because it often changes over time or has patterns that repeat at different times. To make better predictions, a new method called frequency adaptive normalization (FAN) was developed to handle both changing trends and seasonal patterns. This method uses special math to identify the most important repeating patterns in the data, then tries to predict how these patterns will change in the future. The FAN method is designed to work with different types of prediction models and can make predictions more accurate. |
Keywords
» Artificial intelligence » Mse » Time series
