Summary of A Decomposition Modeling Framework For Seasonal Time-series Forecasting, by Yining Pang and Chenghan Li
A Decomposition Modeling Framework for Seasonal Time-Series Forecasting
by Yining Pang, Chenghan Li
First submitted to arxiv on: 12 Dec 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 Multi-scale Seasonal Decomposition Model (MSSD) is introduced for accurate future prediction of seasonal time-series data. The model initially decomposes the univariate time series into three primary components: Ascending, Peak, and Descending, leveraging inherent periodicity to enhance capture of periodic features. A multi-scale network structure is designed to effectively capture various peak fluctuation patterns in the Peak component. Conv2d and Temporal Convolutional Networks are integrated to concurrently capture global and local features, and multi-scale reshaping is incorporated to augment modeling capacity for peak fluctuation patterns. The proposed methodology is validated using three publicly accessible seasonal datasets, achieving a 10% reduction in error compared to baseline models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper develops a new way to predict future values in time series data that changes with the seasons. It breaks down these kinds of time series into smaller parts and creates a model that can capture different patterns and trends. The model uses special techniques, like convolutional neural networks, to analyze the data from different angles. It’s tested on real-world datasets and performs better than existing methods. |
Keywords
* Artificial intelligence * Time series
