Supervised Convolutional Neural Networks for Classification of Flaring and Nonflaring Active Regions Using Line-of-sight Magnetograms

Bhattacharjee, Subhro; Alshehhi, Rasha; Dhuri, Dattaraj B.; Hanasoge, Shravan M.

doi:10.3847/1538-4357/ab9c29

Cited by 22 publications

(14 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We remark the attribution maps obtained by integrated gradients are better in terms of resolution and interpretability than what were used in Bhattacharjee et al (2020) and Yi et al (2021). The occlusion method in Bhattacharjee et al (2020) was shown to highlight the area between the opposite polarities, providing only crude attribution. This is because the size of the occlusion mask is usually chosen to be big enough to cover the informative regions.…”

Section: The Emergence Of Preflare Signatures In the Active Region Ev...mentioning

confidence: 80%

“…Among these tools is the class of attribution methods (e.g., Springenberg et al 2015;Selvaraju et al 2017;Shrikumar et al 2017;Sundararajan et al 2017) that attribute a score to gauge the contribution of each input feature for a given input sample. Attribution methods such as the occlusion method and Grad-CAM have been previously used to interpret CNNs in flare prediction applications (Bhattacharjee et al 2020;Yi et al 2021). In this work, we evaluate additional attribution methods (deconvolution, guided backpropagation, DeepLIFT, and integrated gradients) on the interpretation of CNNs trained to predict flares.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Sun

Bobra

Wang

et al. 2022

ApJ

View full text Add to dashboard Cite

We consider the flare prediction problem that distinguishes flare-imminent active regions that produce an M- or X-class flare in the succeeding 24 hr, from quiet active regions that do not produce any flares within ±24 hr. Using line-of-sight magnetograms and parameters of active regions in two data products covering Solar Cycles 23 and 24, we train and evaluate two deep learning algorithms—a convolutional neural network (CNN) and a long short-term memory (LSTM)—and their stacking ensembles. The decisions of CNN are explained using visual attribution methods. We have the following three main findings. (1) LSTM trained on data from two solar cycles achieves significantly higher true skill scores (TSSs) than that trained on data from a single solar cycle with a confidence level of at least 0.95. (2) On data from Solar Cycle 23, a stacking ensemble that combines predictions from LSTM and CNN using the TSS criterion achieves a significantly higher TSS than the “select-best” strategy with a confidence level of at least 0.95. (3) A visual attribution method called “integrated gradients” is able to attribute the CNN’s predictions of flares to the emerging magnetic flux in the active region. It also reveals a limitation of CNNs as flare prediction methods using line-of-sight magnetograms: it treats the polarity artifact of line-of-sight magnetograms as positive evidence of flares.

show abstract

Section: The Emergence Of Preflare Signatures In the Active Region Ev...mentioning

confidence: 80%

Section: Introductionmentioning

confidence: 99%

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Sun

Bobra

Wang

et al. 2022

ApJ

View full text Add to dashboard Cite

show abstract

“…In solar flare prediction, Bhattacharjee et al (2020) applied the occlusion method and found that CNNs pay attention to polarity inversion regions. Yi et al (2021) applied Grad-CAM to CNNs and found that polarity inversion lines in full-disk MDI and HMI magnetograms are highlighted as an important feature for flare prediction.…”

Section: Interpretation Of Cnnsmentioning

confidence: 99%

“…Springenberg et al 2015;Selvaraju et al 2017;Shrikumar et al 2017;Sundararajan et al 2017) that attribute a score to gauge the contribution of each input feature for a given input sample. Attribution methods such as the occlusion method and Grad-CAM have been previously used to interpret CNNs in flare prediction applications (Bhattacharjee et al 2020;Yi et al 2021). In this work, we evaluate additional attribution methods (Deconvolution, Guided Backpropagation, DeepLIFT, and Integrated Gradients) on the interpretation of CNNs trained to predict flares.…”

Section: Introductionmentioning

confidence: 99%

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Sun,

Bobra,

Wang

et al. 2022

Preprint

View full text Add to dashboard Cite

We consider the flare prediction problem that distinguishes flare-imminent active regions that produce an M-or X-class flare in the future 24 hours, from quiet active regions that do not produce any flare within ±24 hours. Using line-of-sight magnetograms and parameters of active regions in two data products covering Solar Cycle 23 and 24, we train and evaluate two deep learning algorithms-CNN and LSTM-and their stacking ensembles. The decisions of CNN are explained using visual attribution methods. We have the following three main findings. (1) LSTM trained on data from two solar cycles achieves significantly higher True Skill Scores (TSS) than that trained on data from a single solar cycle with a confidence level of at least 0.95. (2) On data from Solar Cycle 23, a stacking ensemble that combines predictions from LSTM and CNN using the TSS criterion achieves significantly higher TSS than the "select-best" strategy with a confidence level of at least 0.95. (3) A visual attribution method called Integrated Gradients is able to attribute the CNN's predictions of flares to the emerging magnetic flux in the active region. It also reveals a limitation of CNN as a flare prediction method using line-of-sight magnetograms: it treats the polarity artifact of line-of-sight magnetograms as positive evidence of flares.

show abstract

“…Recently, the application of supervised machine learning methods, especially deep neural networks (DNNs), to solar flare prediction has been a hot topic, and their successful application in research has been reported (Huang et al 2018;Nishizuka et al 2018;Park et al 2018;Chen et al 2019;Domijan et al 2019;Liu et al 2019;Zheng et al 2019;Bhattacharjee et al 2020;Jiao et al 2020;Li et al 2020;Panos & Kleint 2020;Yi et al 2020). However, there is insufficient discussion on how to develop the methods available to real-time operations in space weather forecasting offices, including the methods for validation and verification of the models.…”

Section: Introductionmentioning

confidence: 99%

Operational solar flare prediction model using Deep Flare Net

Nishizuka

Kubo

Sugiura

et al. 2021

Earth Planets Space

View full text Add to dashboard Cite

We developed an operational solar flare prediction model using deep neural networks, named Deep Flare Net (DeFN). DeFN can issue probabilistic forecasts of solar flares in two categories, such as ≥ M-class and < M-class events or ≥ C-class and < C-class events, occurring in the next 24 h after observations and the maximum class of flares occurring in the next 24 h. DeFN is set to run every 6 h and has been operated since January 2019. The input database of solar observation images taken by the Solar Dynamic Observatory (SDO) is downloaded from the data archive operated by the Joint Science Operations Center (JSOC) of Stanford University. Active regions are automatically detected from magnetograms, and 79 features are extracted from each region nearly in real time using multiwavelength observation data. Flare labels are attached to the feature database, and then, the database is standardized and input into DeFN for prediction. DeFN was pretrained using the datasets obtained from 2010 to 2015. The model was evaluated with the skill score of the true skill statistics (TSS) and achieved predictions with TSS = 0.80 for ≥ M-class flares and TSS = 0.63 for ≥ C-class flares. For comparison, we evaluated the operationally forecast results from January 2019 to June 2020. We found that operational DeFN forecasts achieved TSS = 0.70 (0.84) for ≥ C-class flares with the probability threshold of 50 (40)%, although there were very few M-class flares during this period and we should continue monitoring the results for a longer time. Here, we adopted a chronological split to divide the database into two for training and testing. The chronological split appears suitable for evaluating operational models. Furthermore, we proposed the use of time-series cross-validation. The procedure achieved TSS = 0.70 for ≥ M-class flares and 0.59 for ≥ C-class flares using the datasets obtained from 2010 to 2017. Finally, we discuss the standard evaluation methods for operational forecasting models, such as the preparation of observation, training, and testing datasets, and the selection of verification metrics.

show abstract

Supervised Convolutional Neural Networks for Classification of Flaring and Nonflaring Active Regions Using Line-of-sight Magnetograms

Cited by 22 publications

References 35 publications

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Predicting Solar Flares Using CNN and LSTM on Two Solar Cycles of Active Region Data

Operational solar flare prediction model using Deep Flare Net

Contact Info

Product

Resources

About