Study of Sample Imbalance in Deep Learning Modeling of Solar Flare Forecasting

Solar flares, as violent eruptions occurring in the lower atmosphere of the Sun, exert significant impacts on human activities. Researchers globally have developed multiple prediction models for solar flares, employing empirical, physical, statistical, and other methodologies. There is an order of magnitude difference in the occurrence of different classes of flares. This makes it difficult for traditional convolutional neural network-based flare prediction models to capture M, X class flare features, which leads to the problem of low precision of high level flare prediction. With the breakthrough of deep learning technology in recent years, it has shown strong potential in modelling and prediction of complex problems and a number of works have begun to try to use deep learning methods to construct flare prediction models. In this paper, different deep long-tail learning methods are discussed by us to improve the precision of flare forecasting by controlling the variables for the long-tail distribution phenomenon in flare forecasting. The forecast performance of the model for M and X flares is tried to be improved from the perspectives of training set optimization, loss function optimization and network weight optimization. The experiments on SDO/HMI solar magnetogram data show that the precision of M, X class flare prediction is significantly improved by 53.10% and 38.50%, respectively, and the recall is increased by 64% and 52% compared with the baseline model trained by conventional methods. It shows that the treatment of the long-tailed distribution of data is crucial in the flare forecasting problem, and verifies the effectiveness of the deep long-tailed learning method. This method of improving the precision of tail class forecasts can be applied not only to the field of flare forecasting, but also can be transferred to the analysis of forecasting other typical events of space weather with long-tailed distribution phenomenon.