TY - JOUR
T1 - The Random Hivemind
T2 - An ensemble deep learning application to the solar energetic particle prediction problem
AU - O'Keefe, Patrick M.
AU - Sadykov, Viacheslav
AU - Kosovichev, Alexander
AU - Kitiashvili, Irina N.
AU - Oria, Vincent
AU - Nita, Gelu M.
AU - Francis, Fraila
AU - Chong, Chun Jie
AU - Kosovich, Paul
AU - Ali, Aatiya
AU - Marroquin, Russell D.
N1 - Publisher Copyright:
© 2024 COSPAR
PY - 2024
Y1 - 2024
N2 - The application of machine learning and deep learning techniques, including the wide use of non-ensemble, conventional neural networks (CoNN), for predicting various phenomena has become very popular in recent years thanks to the efficiencies and the abilities of these techniques to find relationships in data without human intervention. However, certain CoNN setups may not work on some datasets, especially if the parameters passed to it, including model parameters and hyperparameters, are arguably arbitrary in nature and need to continuously be updated with the need to retrain the model, especially if the additions of new features render old models obsolete. This concern can be partially alleviated by employing committees of neural networks that are identical in terms of input features and initialized randomly and “vote” on the decisions made by the committees as a whole. Yet, members of the committee have similar architectures and features passed to them, making it possible for the committee members to “agree” on identical sets of weights and biases for all nodes and edges. Members of these committees also cannot be expanded to accommodate new features and entire committees must therefore be retrained in order to do so. We propose the Random Hivemind (RH) approach, which helps to alleviate this concern by having multiple neural network estimators make decisions based on random permutations of features and prescribing a method to determine the weight of the decision of each individual estimator. The effectiveness of RH is demonstrated through experimentation in the predictions of hazardous Solar Energetic Particle (SEP) events by comparing it to that of using both CoNNs and the aforementioned setup of committees identical in input features in this application. Our results demonstrate that RH, while having a comparable or better performance than the CoNN and a Committee-based approach, demonstrates a lesser score spread for the individual experiments, and shows promising results with respect to capturing almost every single flare instance leading to SEPs.
AB - The application of machine learning and deep learning techniques, including the wide use of non-ensemble, conventional neural networks (CoNN), for predicting various phenomena has become very popular in recent years thanks to the efficiencies and the abilities of these techniques to find relationships in data without human intervention. However, certain CoNN setups may not work on some datasets, especially if the parameters passed to it, including model parameters and hyperparameters, are arguably arbitrary in nature and need to continuously be updated with the need to retrain the model, especially if the additions of new features render old models obsolete. This concern can be partially alleviated by employing committees of neural networks that are identical in terms of input features and initialized randomly and “vote” on the decisions made by the committees as a whole. Yet, members of the committee have similar architectures and features passed to them, making it possible for the committee members to “agree” on identical sets of weights and biases for all nodes and edges. Members of these committees also cannot be expanded to accommodate new features and entire committees must therefore be retrained in order to do so. We propose the Random Hivemind (RH) approach, which helps to alleviate this concern by having multiple neural network estimators make decisions based on random permutations of features and prescribing a method to determine the weight of the decision of each individual estimator. The effectiveness of RH is demonstrated through experimentation in the predictions of hazardous Solar Energetic Particle (SEP) events by comparing it to that of using both CoNNs and the aforementioned setup of committees identical in input features in this application. Our results demonstrate that RH, while having a comparable or better performance than the CoNN and a Committee-based approach, demonstrates a lesser score spread for the individual experiments, and shows promising results with respect to capturing almost every single flare instance leading to SEPs.
KW - Computing methodologies: boosting
KW - Machine learning
KW - Neural networks
KW - Solar-terrestrial relations
KW - Sun: activity
KW - Sun: particle emission
UR - http://www.scopus.com/inward/record.url?scp=85194561074&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85194561074&partnerID=8YFLogxK
U2 - 10.1016/j.asr.2024.04.044
DO - 10.1016/j.asr.2024.04.044
M3 - Article
AN - SCOPUS:85194561074
SN - 0273-1177
JO - Advances in Space Research
JF - Advances in Space Research
ER -