Summary of Fan: Fourier Analysis Networks, by Yihong Dong et al.
FAN: Fourier Analysis Networks
by Yihong Dong, Ge Li, Yongding Tao, Xue Jiang, Kechi Zhang, Jia Li, Jinliang Deng, Jing Su, Jun Zhang, Jingjing Xu
First submitted to arxiv on: 3 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
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 proposes FAN, a novel neural network that can effectively model and handle periodic phenomena. Despite the success of general-purpose networks like MLPs and Transformers, they struggle to generalize out-of-domain (OOD) scenarios when dealing with periodic data. To address this limitation, FAN incorporates the Fourier Principle into its structure and computational processes, allowing it to seamlessly replace MLP in various model architectures. Extensive experiments demonstrate the superiority of FAN in periodicity modeling tasks and its effectiveness across a range of real-world tasks, including symbolic formula representation, time series forecasting, language modeling, and image recognition. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary FAN is a new kind of neural network that can understand patterns that repeat over time or space. Right now, computers have trouble recognizing these patterns, which makes it hard for them to do things like predict the weather or recognize music. The researchers who made FAN wanted to change this by creating a network that’s good at recognizing patterns that repeat. They did this by adding something called the Fourier Principle to the way FAN works. This means that FAN can be used in all sorts of situations, from predicting stock prices to recognizing faces. |
Keywords
» Artificial intelligence » Neural network » Time series
