Hongfeng Chen
Hongfeng Chen has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Pengyu Wang
Institute of Network Science and Big Data Technology, Software College, Northeastern University, Shenyang, China
4 shared publications - Tao Ren
Software College, Northeastern University, Shenyang, China
4 shared publications Fanchun Meng
3 shared publicationsTao Ren
3 shared publicationsYanlu Ma
China Earthquake Networks Center, Beijing, China
3 shared publications- Fanchun Meng
Software College, Northeastern University, Shenyang, China
2 shared publications
Latest Publications
- Open Access
- 14 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 68453-68462
A Ghost-Attention Network for Discriminating Tectonic and Non-Tectonic Events on a Small and Imbalanced Dataset
Discrimination between tectonic and non-tectonic events is crucial to assess seismic hazards and manage associated risks. However, the discrimination process is challenging due to the imbalanced distribution of tectonic and non-tectonic events. In this paper, we propose a ghost-attention network (GA-Net) consisting of multiple ghost modules and convolutional block attention modules (CBAMs) to solve this problem. Ghost module allows the network to extract feature maps using cost effective operations, which are suitable for small and unbalanced training sets. CBAM emphasizes meaningful features along channel and spatial axes, effectively learning which information to emphasize or suppress. We train the proposed GA-Net using seismic data from Shaanxi Province in China. The evaluation shows that on the test set, the proposed GA-Net achieves sensitivity, specificity, and accuracy of 96.40%, 93.24%, and 95.75%, respectively. In addition, a comparative analysis with several state-of-the-art networks establishes its superiority. Our GA-Net exhibits robust generalized performance on several training sets with different imbalanced levels and an independent dataset.
- 14 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 68453-68462
Open AccessA Ghost-Attention Network for Discriminating Tectonic and Non-Tectonic Events on a Small and Imbalanced Dataset
Discrimination between tectonic and non-tectonic events is crucial to assess seismic hazards and manage associated risks. However, the discrimination process is challenging due to the imbalanced distribution of tectonic and non-tectonic events. In this paper, we propose a ghost-attention network (GA-Net) consisting of multiple ghost modules and convolutional block attention modules (CBAMs) to solve this problem. Ghost module allows the network to extract feature maps using cost effective operations, which are suitable for small and unbalanced training sets. CBAM emphasizes meaningful features along channel and spatial axes, effectively learning which information to emphasize or suppress. We train the proposed GA-Net using seismic data from Shaanxi Province in China. The evaluation shows that on the test set, the proposed GA-Net achieves sensitivity, specificity, and accuracy of 96.40%, 93.24%, and 95.75%, respectively. In addition, a comparative analysis with several state-of-the-art networks establishes its superiority. Our GA-Net exhibits robust generalized performance on several training sets with different imbalanced levels and an independent dataset.
- 13 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 62 (01962892) , 1-11
SeisParaNet: A Novel Multitask Network for Seismic Source Characterization in Earthquake Early Warning
Rapid and accurate seismic source characterization significantly influences the performance of Earthquake Early Warning (EEW) systems. However, the complexity of the seismic source modeling and the error accumulation during continuous characterization make it difficult to accurately characterize various source parameters. Furthermore, current artificial intelligence methods focus on a single task, lacking inter-task fusion and guidance from specialized knowledge. In this study, we propose a novel Deep Learning (DL) algorithm (SeisParaNet) to estimate P-wave arrival time, source location, and magnitude simultaneously based on a multi-task framework. To exploit seismological knowledge and attenuate the strong inter-task dependencies, this study incorporates arrival time differences information into the analysis of source localization parameters by using the attention mechanism and incorporates source location features into estimating local magnitude (ML). In addition, SeisParaNet uses a probability-based Self-Attention mechanism (Prob-Attention) to extract temporal information from waveforms. Experimental results demonstrate that, following a limited number of trainings on the STanford EArthquake Dataset (STEAD), SeisParaNet exhibits the capability to capture complex seismic patterns and rapidly characterize seismic sources. Furthermore, the introduction of Prob-Attention reduces computational complexity by 67%, validating the potential of SeisParaNet in EEW applications.
- 13 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 62 (01962892) , 1-11
SeisParaNet: A Novel Multitask Network for Seismic Source Characterization in Earthquake Early Warning
Rapid and accurate seismic source characterization significantly influences the performance of Earthquake Early Warning (EEW) systems. However, the complexity of the seismic source modeling and the error accumulation during continuous characterization make it difficult to accurately characterize various source parameters. Furthermore, current artificial intelligence methods focus on a single task, lacking inter-task fusion and guidance from specialized knowledge. In this study, we propose a novel Deep Learning (DL) algorithm (SeisParaNet) to estimate P-wave arrival time, source location, and magnitude simultaneously based on a multi-task framework. To exploit seismological knowledge and attenuate the strong inter-task dependencies, this study incorporates arrival time differences information into the analysis of source localization parameters by using the attention mechanism and incorporates source location features into estimating local magnitude (ML). In addition, SeisParaNet uses a probability-based Self-Attention mechanism (Prob-Attention) to extract temporal information from waveforms. Experimental results demonstrate that, following a limited number of trainings on the STanford EArthquake Dataset (STEAD), SeisParaNet exhibits the capability to capture complex seismic patterns and rapidly characterize seismic sources. Furthermore, the introduction of Prob-Attention reduces computational complexity by 67%, validating the potential of SeisParaNet in EEW applications.
- 2 January 2024
- Published by Elsevier in Computers & Geosciences
- Vol. 184
Determination of earthquake focal mechanism via multi-task learning
- 2 January 2024
- Published by Elsevier in Computers & Geosciences
- Vol. 184
Determination of earthquake focal mechanism via multi-task learning
- 28 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 61 (01962892) , 1-13
BTRPP: A Rapid PGA Prediction Model Based on Machine Learning
Peak ground acceleration (PGA) is a critical parameter in the postearthquake intensity snapshot. In this article, we propose a Bagged Tree for Rapid PGA Prediction (BTRPP) based on velocity waveforms recorded at stations around the epicenter, as well as a feature pruning algorithm based on BTRPP to predict PGA and analyze the relationship between PGA and manually predefined features. The research not only aims to discover the best model for forecasting PGA but also examines the impact of features using the interpretability of some models. As inputs to the model, BTRPP uses and mixes features in the time and frequency domains. Bayesian optimization (BO) is used to find the optimum hyperparameters of BTRPP. The results demonstrate that, using the data within 5 s after the first arrival of the P-wave, the BTRPP with BO can reduce the RMSE to 0.2076 and enhance the R-square to 0.91. It is, finally, found that the PGA is most closely related to the average positive amplitude, average negative amplitude, and spectral slope through the feature pruning algorithm in this article, which indicates that the results are plausible.
- 28 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 61 (01962892) , 1-13
BTRPP: A Rapid PGA Prediction Model Based on Machine Learning
Peak ground acceleration (PGA) is a critical parameter in the postearthquake intensity snapshot. In this article, we propose a Bagged Tree for Rapid PGA Prediction (BTRPP) based on velocity waveforms recorded at stations around the epicenter, as well as a feature pruning algorithm based on BTRPP to predict PGA and analyze the relationship between PGA and manually predefined features. The research not only aims to discover the best model for forecasting PGA but also examines the impact of features using the interpretability of some models. As inputs to the model, BTRPP uses and mixes features in the time and frequency domains. Bayesian optimization (BO) is used to find the optimum hyperparameters of BTRPP. The results demonstrate that, using the data within 5 s after the first arrival of the P-wave, the BTRPP with BO can reduce the RMSE to 0.2076 and enhance the R-square to 0.91. It is, finally, found that the PGA is most closely related to the average positive amplitude, average negative amplitude, and spectral slope through the feature pruning algorithm in this article, which indicates that the results are plausible.
- 24 March 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 234 (2) , 1355-1362
Seismic severity estimation using convolutional neural network for earthquake early warning
In this study, magnitude estimation in earthquake early warning (EEW) systems is seen as a classification problem: the single-channel waveform, starting from the P-wave onset and lasting 4 s, is given in the input, and earthquake severity (medium and large earthquakes: local magnitude (ML) ≥ 5; small earthquakes: ML < 5) is the classification result. The convolutional neural network (CNN) is proposed to estimate the severity of the earthquake, which is composed of several blocks that can extract the latent representation of the input from different receptive fields automatically. We train and test the proposed CNN model using two data sets. One is recorded by the China Earthquake Networks Center (CENC), and the other is the Stanford Earthquake Dataset (STEAD). Accordingly, the proposed CNN model achieves a test accuracy of 97.90 per cent. The proposed CNN model is applied to estimate two real-world earthquake swarms in China (the Changning earthquake and the Tangshan earthquake swarms) and the INSTANCE data set, and demonstrated the promising performance of generalization. In addition, the proposed CNN model has been connected to the CENC for further testing using real-world real-time seismic data.
- 24 March 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 234 (2) , 1355-1362
Seismic severity estimation using convolutional neural network for earthquake early warning
In this study, magnitude estimation in earthquake early warning (EEW) systems is seen as a classification problem: the single-channel waveform, starting from the P-wave onset and lasting 4 s, is given in the input, and earthquake severity (medium and large earthquakes: local magnitude (ML) ≥ 5; small earthquakes: ML < 5) is the classification result. The convolutional neural network (CNN) is proposed to estimate the severity of the earthquake, which is composed of several blocks that can extract the latent representation of the input from different receptive fields automatically. We train and test the proposed CNN model using two data sets. One is recorded by the China Earthquake Networks Center (CENC), and the other is the Stanford Earthquake Dataset (STEAD). Accordingly, the proposed CNN model achieves a test accuracy of 97.90 per cent. The proposed CNN model is applied to estimate two real-world earthquake swarms in China (the Changning earthquake and the Tangshan earthquake swarms) and the INSTANCE data set, and demonstrated the promising performance of generalization. In addition, the proposed CNN model has been connected to the CENC for further testing using real-world real-time seismic data.
- 23 February 2023
- Published by Elsevier in Computers & Geosciences
- Vol. 173
NT-Com: A combined machine learning model for picking up first arrival
- 23 February 2023
- Published by Elsevier in Computers & Geosciences
- Vol. 173
NT-Com: A combined machine learning model for picking up first arrival
- 12 June 2019
- Published by Springer Nature in Pure and Applied Geophysics
- Vol. 177 (1) , 455-467
Classification of Tectonic and Nontectonic Earthquakes by an Integrated Learning Algorithm
- 12 June 2019
- Published by Springer Nature in Pure and Applied Geophysics
- Vol. 177 (1) , 455-467
Classification of Tectonic and Nontectonic Earthquakes by an Integrated Learning Algorithm