Summary of Large Scale Evaluation Of Deep Learning-based Explainable Solar Flare Forecasting Models with Attribution-based Proximity Analysis, by Temitope Adeyeha et al.
Large Scale Evaluation of Deep Learning-based Explainable Solar Flare Forecasting Models with Attribution-based Proximity Analysis
by Temitope Adeyeha, Chetraj Pandey, Berkay Aydin
First submitted to arxiv on: 27 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Solar and Stellar Astrophysics (astro-ph.SR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
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 Deep learning models have shown promise in predicting solar flares, but current evaluations focus on accuracy while neglecting interpretability and reliability. To address this gap, a novel proximity-based framework is proposed for analyzing post hoc explanations to assess the interpretability of deep learning models for solar flare prediction. Two models trained on full-disk line-of-sight magnetogram images are compared, using the Guided Gradient-weighted Class Activation Mapping method to generate attribution maps and analyze decision-making processes. A proximity-based metric is introduced to evaluate explanation accuracy and relevance in operational systems, revealing that model predictions align with active region characteristics to varying degrees. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Solar flares can damage Earth’s infrastructure, so predicting them accurately is crucial. Researchers have used deep learning models to predict solar flares, but they haven’t looked at how these models work or why they make certain decisions. To fix this, a new way of analyzing the models’ explanations was developed. This method helps understand how the models work and what they’re looking for when predicting solar flares. The results show that the models are good at predicting solar flares based on active regions. |
Keywords
* Artificial intelligence * Deep learning
