Summary of Time Is Not Enough: Time-frequency Based Explanation For Time-series Black-box Models, by Hyunseung Chung et al.
Time is Not Enough: Time-Frequency based Explanation for Time-Series Black-Box Models
by Hyunseung Chung, Sumin Jo, Yeonsu Kwon, Edward Choi
First submitted to arxiv on: 7 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Signal Processing (eess.SP)
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 paper proposes Spectral eXplanation (SpectralX), an XAI framework that provides time-frequency explanations for time-series black-box classifiers. It also introduces Feature Importance Approximations (FIA), a new perturbation-based XAI method that enhances computational efficiency and class-specific explanations in time-series classification tasks. The authors compare the explanation performance of FIA with other existing methods in both time-domain and time-frequency domain, showing the superiority of their approach in the latter. A user study confirms the practicality of FIA in the SpectralX framework for class-specific time-frequency based time-series explanations. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a new way to understand how computer models work when they’re predicting things that happen over time. It’s like trying to figure out why a model thought a certain day would be sunny or rainy, but instead of looking at just the temperature and weather patterns, it also looks at other things that might affect the outcome, like the time of year or the weather in previous days. The new method is called SpectralX, and it’s better than other methods because it can understand both the time and frequency features of the data. It also includes a way to explain why certain features are important for making predictions. |
Keywords
» Artificial intelligence » Classification » Temperature » Time series
