{ "fetched_at": "2026-10-19T21:58:01Z", "source_url": "https://www.kaggle.com/competitions/bigdata2020-flare-prediction/data", "pages": [ { "name": "rules", "content": "This is a page where you can include rules that participants must accept before joining. You may wish to include rules like:\n\n - Don't cheat!\n - Apply yourself!\n - Have fun!", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "data-description", "content": "Download the entire dataset from [[https://dmlab.cs.gsu.edu/solar/data/data-comp-2020/](https://dmlab.cs.gsu.edu/solar/data/data-comp-2020/)]\n\n----\n\n### Flare Classes\n\nSolar flares are automatically detected and classified by the National Oceanic and Atmospheric Administration’s (NOAA) constellation of GOES satellites based on their peak flux in soft X-ray wavelengths on a logarithmic scale as A-, B- ,C- , M-, and X-class solar flares, from weakest to the strongest. The most intense of these classes, namely M and X, are most often targeted for prediction due to their potentially adverse space-weather ramifications. In this competition, the dataset consists of 5 classes, namely X, M, C, B, and FQ. The last one represents the flare-quiet regions where no flare has been detected within the observation period. Note that we do not include A class flares for the reason given above.\n\n\n### Dataset\n\nThe data in each of the following files are formatted as sets of json objects. Each one containing a unique identifier within the file, a class label if one of the training folds, and an array of time series values used for prediction. Each sample multi-variate time series has 60 values representing an observation taken at 12-minute intervals. The first value in the array is the sample observation taken at the furthest point in time from the predication period and the last value in the array is the sample observation taken at the closest point in time to the prediction period. The period of observations represented by each labeled time series is a 12 hour window of observations sliced from a longer time series.\n\nSince the dataset is a set of partially overlapping sliced samples pulled from larger multi-variate time series, sampling decisions can have deleterious effects to the generalization of models trained using different strategies. For this reason, each of the training folds are temporally disjoint sets of samples so as to mitigate any possible over-fitting issues that may arise from sampling decisions made by the participants. If a standard ⅔ training ⅓ testing sampling strategy is utilized to determine a model for use as your final submission (i.e. selecting 2 folds for training and 1 fold for testing), it is guaranteed that no partially overlapping samples will be selected for both the training and testing sets and therefore the resultant trained model is more likely to generalize to the final test set used for scoring. This is not to say that using all three training folds for training your final model will not be beneficial, it is simply providing a way for participants to train/test their models independently from having to submit a result set for evaluation (which teams are limited to 2 per day), and avoid common over-fitting errors. \n\n### File Description\n* **Training Sets:**\n * `train_partition1_data.json`\n * `train_partition2_data.json`\n * `train_partition3_data.json`\n* **Testing Set:**\n * `test_4_5_data.json`\n* **Sample Submission**:\n * A sample submission file in the correct format: `sampleSubmission.csv`\n\n### Data Structures\n* In training data: `{: {\"label\": , \"values\": }}`\n* In testing data: `{: {\"label\": none, \"values\": }}`\n\nwhere `` is a dictionary representing the mvts.\n\nFor more details, see `data_demo` notebook in the Notebooks page.\n----------", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Introduction", "content": "

Introduction

\n

This competition is part of the Solar Flare Prediction from Time Series of Solar Magnetic Field Parameters in the IEEE Big Data 2020 big Data Cup.

\n\n

\nAfter the competition phase is completed, a link for the submission of the accompanying academic paper will be provided to the top 10 participants as ranked by the public/private leaderboard weighting described on the competition website. The academic papers will be ranked by peer reviewers and a final decision will be made using the weighting method detailed on the competition website.\n

\n\n

Important Notes

\n

\n

  • We are interested in the ideas and their implementations, and not necessarily the highest score. Therefore, we strongly encourage contestants to keep their code in a (for now, private) repository and share it with us.
    • \n
  • There are other evaluation metrics (e.g., `TSS` and `HSS`) that are more appropriate for rare-event classification tasks such as ours. Since they are not available at Kaggle, for ranking the ideas we rely on the most relevant metric that works for our multi-class task. But in the academic paper, if you are interested in continuing in that direction, we certainly need other metrics to be included as well. We would be more than happy to provide more information on this topic.
    • \n

    \n\n

    Organizers

    \n

    \nOn behalf of the Data Mining Lab ( dmlab.cs.gsu.edu ), in the Department of Computer Science, Georgia State University, this challenge has been organized by the following.\n\n

    \n\n
  • Dustin J. Kempton, Ph.D. (cs.gsu.edu/~dkempton1)\n
    • \n\n
  • Berkay Aydin, Ph.D. (cs.gsu.edu/~baydin/)\n
    • \n\n
  • Azim Ahmadzadeh, Ph.D. Candidate (grid.cs.gsu.edu/~aahmadzadeh1/)\n
    • \n\n
  • Rafal A. Angryk, Ph.D. (cs.gsu.edu/~rangryk)\n
    • \n\n

    ", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Evaluation", "content": "

    The evaluation metric for this competition is FBeta-Score with Beta set to 1, which makes the score equal to the F1 score. The F1 score, commonly used in information retrieval, measures accuracy using the statistics precision p and recall r. Precision is the ratio of true positives (tp) to all predicted positives (tp + fp). Recall is the ratio of true positives to all actual positives (tp + fn). The F1 score is given by:

    \n

    \\[ F1 = 2\\frac{p \\cdot r}{p+r}\\ \\ \\mathrm{where}\\ \\ p = \\frac{tp}{tp+fp},\\ \\ r = \\frac{tp}{tp+fn} \\]

    \n

    The F1 metric weights recall and precision equally, and a good retrieval algorithm will maximize both precision and recall simultaneously. Thus, moderately good performance on both will be favored over extremely good performance on one and poor performance on the other.

    \n\n

    \nSince we are dealing with multi-class classification problem, we use `MeanFScore` metric of Kaggle, which is equivalent to `sklearn.metrics.f1_score(y_true, y_pred, average='micro')`. For more information, see [sklearn's documentation](https://scikit-learn.org/stable/modules/generated/sklearn.metrics.fbeta_score.html).\n<\\p>\n\n

    Submission Format

    \n

    For every time series sample in the testSet.json file, submission files should contain two columns: 'Id' and 'Label'. The 'Label' is the class membership of each of the samples. The results are evaluated based on the 'Id' and 'Label' pairs. So, ordering the results by 'Id' is not necessary.

    \n\n

    The file should contain a header and have the following format:

    \n
    Id,Label
    1,X
    ", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Timeline", "content": "

    Important Dates

    \n\n
  • May 8, 2020: Competition Opens (team submission is open to any interested individual(s))
    • \n\n
  • September 1, 2020: Phase 1 Ends: Result submission used for ranking closes.
    • \n\n
  • September 2, 2020: Phase 2 Begins: Top participants invited to submit an academic paper.
    • \n\n
  • October 1, 2020: Phase 2 Ends: Academic paper submissions due.
    • \n\n
  • November 1, 2020: Notification of paper acceptance.
    • \n\n
  • November 15, 2020: Camera-ready of accepted papers.
    • \n\n
  • December 10-13, 2020 (TBD): Conference (Part of [SABiD Workshop](https://dmlab.cs.gsu.edu/bigdata/flare-comp-2020/) (Dec 10 or 13 TBD))
    • \n", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Description", "content": "

    Solar flares and Coronal Mass Ejections (CMEs) [1, 2, 3] are events occurring in the solar corona and heliosphere that can have a major negative impact on our technology dependent society [4]. Electromagnetic radiation and ionized particles from solar flares and eruptions tend to be filtered out by Earth’s atmosphere, but they can still pose a hazard to astronauts and sensitive equipment in space, as well as disrupt various high frequency radio communications that military and civilian customers become increasingly reliant upon each year. A strong enough CME can also cause significant enough fluctuations in Earth’s magnetosphere to induce currents in large networks of conductive materials such as power grids. These induced currents can lead to surges that have the potential to melt transformers of long distance transmission lines causing large scale blackouts. A 2008 report by the National Research Council concluded that a solar superstorm, similar to one observed in 1857 called the Carrington event [5], could cripple the entire US power grid for months and lead to an economic damage of 1 to 2 trillion dollars [6].

    \n

    \nIn response, the White House released the National Space Weather Strategy and Space Weather Action Plan [4] in 2015 as a roadmap for producing research that focuses on predicting and mitigating the effects of solar eruptive activity. One of the suggested routes to accomplishing the goals set out by this roadmap was to use machine learning to predict extreme space weather events, which was reiterated recently by the current administration [7]. The goal of this dataset competition is to introduce the machine learning/data mining community to an integrated dataset that can be utilized for predicting and understanding solar flares.

    \n\n

    Acknowledgements

    \n

    This work was supported in part by two NASA Grant Awards [No. NNX11AM13A, and No. NNX15AF39G], and one NSF Grant Award [No. AC1443061]. The NSF Grant Award has been supported by funding from the Division of Advanced Cyberinfrastructure within the Directorate for Computer and Information Science and Engineering, the Division of Astronomical Sciences within the Directorate for Mathematical and Physical Sciences, and the Division of Atmospheric and Geospace Sciences within the Directorate for Geosciences. Also, we would like to mention that all images used in this work are courtesy of NASA/SDO and the AIA, EVE, and HMI science teams.

    \n\n\n

    References

    \n\n 1. Benz, A. O. Flare observations. Living Reviews in Solar Physics 5, 1 (2008). URL https://doi.org/10.12942/lrsp-2008-1.\n 2. Howard, T. Coronal mass ejections: An introduction, vol. 376 (Springer Science |& Business Media, 2011).\n 3. Martens, P. C. & Angryk, R. A. Data Handling and Assimilation for Solar Event Prediction. In Foullon, C. & Malandraki, O. E. (eds.) Space Weather of the Heliosphere: Processes and Forecasts, vol. 335 of IAU Symposium, 344–347 (2018). \"[arXiv:1712.01402][1]\".\n 4. National Science and Technology Council. National space weather action plan (2015). Available at https://obamawhitehouse.archives.gov/sites/default/files/microsites/ostp/final_nationalspaceweatheractionplan_20151028.pdf.\n 5. Carrington, R. C. Description of a singular appearance seen in the sun on september 1, 1859. Monthly Notices of the Royal Astronomical Society 20, 13–15 (1859).\n 6. National Research Council. Severe space weather events: Understanding societal and economic impacts: A workshop report (National Academies Press, 2009).\n 7. Hutson, M. Trump to launch artificial intelligence initiative, but many details lacking. Science (2019). URL https://doi.org/10.1126/science.aaw9677.\n\n\n [1]: https://arxiv.org/abs/1712.01402", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false } ] }