A5101582874- Solar Radiation and Photovoltaics
- Solar and Space Plasma Dynamics
- Earthquake Detection and Analysis
- Gut microbiota and health
- Computational Physics and Python Applications
- Bioinformatics and Genomic Networks
- Gaussian Processes and Bayesian Inference
- Tensor decomposition and applications
- Geomagnetism and Paleomagnetism Studies
- Metabolomics and Mass Spectrometry Studies
- Machine Learning and Data Classification
- Anomaly Detection Techniques and Applications
- Scientific Research and Discoveries
- Photovoltaic System Optimization Techniques
- Statistical and numerical algorithms
- Neural Networks and Applications
Griffith University
2025
University of Michigan
2019-2022
Michigan United
2022
We present several methods towards construction of precursors, which show great promise early predictions, solar flare events in this paper. A data pre-processing pipeline is built to extract useful from multiple sources, Geostationary Operational Environmental Satellites (GOES) and Solar Dynamics Observatory (SDO)/Helioseismic Magnetic Imager (HMI), prepare inputs for machine learning algorithms. Two classification models are presented: flares quiet times active regions strong versus weak...
Abstract We consider the flare prediction problem that distinguishes flare-imminent active regions produce an M- or X-class in succeeding 24 hr, from quiet do not any flares within ±24 hr. Using line-of-sight magnetograms and parameters of two data products covering Solar Cycles 23 24, we train evaluate deep learning algorithms—a convolutional neural network (CNN) a long short-term memory (LSTM)—and their stacking ensembles. The decisions CNN are explained using visual attribution methods....
Predicting the dynamics and functions of microbiomes constructed from bottom-up is a key challenge in exploiting them to our benefit. Current models based on ecological theory fail capture complex community behaviors due higher order interactions, do not scale well with increasing complexity considering multiple functions. We develop apply long short-term memory (LSTM) framework advance understanding assembly health-relevant metabolite production using synthetic human gut community. A...
The China Brain Multi-omics Atlas Project (CBMAP) aims to generate a comprehensive molecular reference map of over 1,000 human brains (Phase I), spanning broad age range and multiple regions in China, address the underrepresentation East Asian populations brain research. By integrating genome, epigenome, transcriptome, proteome (including post-translational modifications), metabolome data, CBMAP is set provide rich invaluable resource for investigating underpinnings aging-related phenotypes...
Abstract Predicting the dynamics and functions of microbiomes constructed from bottom-up is a key challenge in exploiting them to our benefit. Current ordinary differential equation-based models fail capture complex behaviors that fall outside predetermined ecological theory do not scale well with increasing community complexity considering multiple functions. We develop apply long short-term memory (LSTM) framework advance understanding assembly health-relevant metabolite production using...
In this paper, we consider incorporating data associated with the sun’s north and south polar field strengths to improve solar flare prediction performance using machine learning models. When used supplement local from active regions on photospheric magnetic of sun, provides global information predictor. While such features have been previously proposed for predicting next cycle’s intensity, in paper propose them help classify individual flares. We conduct experiments HMI employing four...
Many applications produce multiway data of exceedingly high dimension. Modeling such multi-way is important in multichannel signal and video processing where sensors multi-indexed data, e.g. over spatial, frequency, temporal dimensions. We will address the challenges covariance representation review some progress statistical modeling past two decades, focusing on tensor-valued models their inference. illustrate through a space weather application: predicting evolution solar active regions time.
Earth and Space Science Open Archive This is a preprint has not been peer reviewed. ESSOAr venue for early communication or feedback before review. Data may be preliminary.Learn more about preprints preprintOpen AccessYou are viewing the latest version by default [v3]Predicting Solar Flares using CNN LSTM on Two Cycles of Active Region DataAuthorsZeyuSuniDMonicaBobraiDXiantongWangiDYuWangiDHuSunTamasGombosiiDYangCheniDAlfredHeroiDSee all authors Zeyu SuniDCorresponding Author• Submitting...
Earth and Space Science Open Archive This is a preprint has not been peer reviewed. ESSOAr venue for early communication or feedback before review. Data may be preliminary.Learn more about preprints preprintOpen AccessYou are viewing an older version [v2]Go to new versionPredicting Solar Flares using CNN LSTM on Two Cycles of Active Region DataAuthorsZeyuSuniDMonicaBobraiDXiantongWangiDYuWangiDHuSunTamasGombosiiDYangCheniDAlfredHeroiDSee all authors Zeyu SuniDCorresponding Author• Submitting...
Earth and Space Science Open Archive This is a preprint has not been peer reviewed. ESSOAr venue for early communication or feedback before review. Data may be preliminary.Learn more about preprints preprintOpen AccessYou are viewing an older version [v1]Go to new versionPredicting Solar Flares using CNN LSTM on Two Cycles of Active Region DataAuthorsZeyuSuniDMonicaBobraiDXiantongWangiDYuWangiDHuSunTamasGombosiiDYangCheniDAlfredHeroiDSee all authors Zeyu SuniDCorresponding Author• Submitting...
In this paper we present several methods to identify precursors that show great promise for early predictions of solar flare events.A data preprocessing pipeline is built extract useful from multiple sources, Geostationary Operational Environmental Satellites and Solar Dynamics Observatory (SDO)/Helioseismic Magnetic Imager (HMI), prepare inputs machine learning algorithms.Two classification models are presented: flares quiet times active regions strong versus weak events.We adopt deep...
The increasing availability of high dimensional multi-indexed data has created an opportunity for the development graphical modeling software that bridges gap between second order statistical and computational models. In this paper, we introduce TensorGraphicalModels, a suite Julia tools estimation high-dimensional multiway (tensor-variate) covariance precision matrices, with applications to ensemble Kalman filtering. package implements several state-of-the-art matrix estimators. These are...