Summary of Correlating Time Series with Interpretable Convolutional Kernels, by Xinyu Chen et al.
Correlating Time Series with Interpretable Convolutional Kernels
by Xinyu Chen, HanQin Cai, Fuqiang Liu, Jinhua Zhao
First submitted to arxiv on: 2 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 This paper addresses the challenge of convolutional kernel learning in univariate, multivariate, and multidimensional time series data. It proposes a sparse regression approach for univariate time series, leveraging circular convolution and circulant matrices. The method is then generalized to multivariate and multidimensional data using tensor computations and standard sparse regression techniques. The optimization problem is solved using the non-negative subspace pursuit method, allowing the convolutional kernel to capture temporal patterns and correlations. The proposed model is evaluated on real-world time series datasets, including rideshare and taxi trip data from New York City and Chicago, which reveals interpretable local correlations and cyclical patterns. In multidimensional fluid flow data, the convolutional kernels can reinforce tensor factorization, improving performance in reconstruction tasks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps machines learn from time series data, like temperature readings or stock prices. It shows how to find patterns in this kind of data by treating it as a special type of math problem. The method works on one-dimensional, two-dimensional, and three-dimensional data, which is useful for things like predicting weather patterns or analyzing traffic flow. The paper also tests its approach on real-world datasets, including ride-sharing data from New York City and Chicago, where it finds weekly seasonal patterns. |
Keywords
» Artificial intelligence » Optimization » Regression » Temperature » Time series
