Search for: All records

Abstract Solar energetic particle (SEP) events pose significant risks to both space and ground-level infrastructure, as well as to human health in space. Understanding and predicting these events are critical for mitigating their potential impacts. In this paper, we address the challenge of predicting SEP events using proton flux data. We leverage some of the most recent advances in time series data mining, such as shapelets and the matrix profile, to propose a simple and easily understandable prediction approach. Our objective is to mitigate the interpretability challenges inherent to most machine learning models and to show that other methods exist that can not only yield accurate forecasts but also facilitate exploration and insight generation within the data domain. For this purpose, we construct a multivariate time series data set consisting of proton flux data recorded by the National Oceanic and Atmospheric Administration's geosynchronous orbit Earth-observing satellite. Then, we use our proposed approach to mine shapelets and make predictions using a random forest classifier. We demonstrate that our approach rivals state-of-the-art SEP prediction, offering superior interpretability and the ability to predict SEP events before their parent eruptive flares. 

Solar wind modeling is classified into two main types: empirical models and physics-based models, each designed to forecast solar wind properties in various regions of the heliosphere. Empirical models, which are cost-effective, have demonstrated significant accuracy in predicting solar wind at the L1 Lagrange point. On the other hand, physics-based models rely on magnetohydrodynamics (MHD) principles and demand more computational resources. In this research paper, we build upon our recent novel approach that merges empirical and physics-based models. Our recent proposal involves the creation of a new physics-informed neural network that leverages time series data from solar wind predictors to enhance solar wind prediction. This innovative method aims to combine the strengths of both modeling approaches to achieve more accurate and efficient solar wind predictions. In this work, we show the variability of the proposed physics-informed loss across multiple deep learning models. We also study the effect of training the models on different solar cycles on the model’s performance. This work represents the first effort to predict solar wind by integrating deep learning approaches with physics constraints and analyzing the results across three solar cycles. Our findings demonstrate the superiority of our physics-constrained model over other unconstrained deep learning predictive models. 

In this paper, we propose Attention-based Counterfactual Explanation (AB-CF), a novel model that generates post-hoc counterfactual explanations for multivariate time series classifcation that narrow the attention to a few important segments. We validated our model using seven real-world time-series datasets from the UEA repository. Our experimental results show the superiority of ABCF in terms of validity, proximity, sparsity, contiguity, and effciency compared with other competing state-of-the-art baselines.