Summary of Fpn-fusion: Enhanced Linear Complexity Time Series Forecasting Model, by Chu Li et al.
FPN-fusion: Enhanced Linear Complexity Time Series Forecasting Model
by Chu Li, Pingjia Xiao, Qiping Yuan
First submitted to arxiv on: 6 Jun 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 FPN-fusion model demonstrates superior predictive performance for time series prediction tasks while maintaining linear computational complexity. The model incorporates a Feature Pyramid Network (FPN) to capture data characteristics and a multi-level fusion structure to integrate deep and shallow features. Empirically, FPN-fusion outperforms DLiner in 31 out of 32 test cases on eight open-source datasets, achieving an average reduction of 16.8% in mean squared error (MSE) and 11.8% in mean absolute error (MAE). Additionally, compared to the transformer-based PatchTST, FPN-fusion achieves 10 best MSE and 15 best MAE results while using only 8% of PatchTST’s total computational load. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a new way to predict things that happen over time. They make a model called FPN-fusion that is really good at doing this, even when it doesn’t use as much computer power. The model does two special things: it looks at the data in different ways (like looking at big and small patterns), and it combines these views to get an even better result. This new way works well on lots of different datasets, beating some other models that are already good. It’s like having a superpower for predicting what might happen next! |
Keywords
» Artificial intelligence » Feature pyramid » Mae » Mse » Time series » Transformer
