Attention-Based Bi-Directional Long-Short Term Memory Network for Earthquake Prediction. Banna, M. H. A., Ghosh, T., Nahian, M. J. A., Taher, K. A., Kaiser, M. S., Mahmud, M., Hossain, M. S., & Andersson, K. IEEE Access, 9:56589–56603, 2021. Conference Name: IEEE Accessdoi abstract bibtex 2 downloads An earthquake is a tremor felt on the surface of the earth created by the movement of the major pieces of its outer shell. Till now, many attempts have been made to forecast earthquakes, which saw some success, but these attempted models are specific to a region. In this paper, an earthquake occurrence and location prediction model is proposed. After reviewing the literature, long short-term memory (LSTM) is found to be a good option for building the model because of its memory-keeping ability. Using the Keras tuner, the best model was selected from candidate models, which are composed of combinations of various LSTM architectures and dense layers. This selected model used seismic indicators from the earthquake catalog of Bangladesh as features to predict earthquakes of the following month. Attention mechanism was added to the LSTM architecture to improve the model’s earthquake occurrence prediction accuracy, which was 74.67%. Additionally, a regression model was built using LSTM and dense layers to predict the earthquake epicenter as a distance from a predefined location, which provided a root mean square error of 1.25.
@article{banna_attention-based_2021,
title = {Attention-{Based} {Bi}-{Directional} {Long}-{Short} {Term} {Memory} {Network} for {Earthquake} {Prediction}},
volume = {9},
issn = {2169-3536},
doi = {10.1109/ACCESS.2021.3071400},
abstract = {An earthquake is a tremor felt on the surface of the earth created by the movement of the major pieces of its outer shell. Till now, many attempts have been made to forecast earthquakes, which saw some success, but these attempted models are specific to a region. In this paper, an earthquake occurrence and location prediction model is proposed. After reviewing the literature, long short-term memory (LSTM) is found to be a good option for building the model because of its memory-keeping ability. Using the Keras tuner, the best model was selected from candidate models, which are composed of combinations of various LSTM architectures and dense layers. This selected model used seismic indicators from the earthquake catalog of Bangladesh as features to predict earthquakes of the following month. Attention mechanism was added to the LSTM architecture to improve the model’s earthquake occurrence prediction accuracy, which was 74.67\%. Additionally, a regression model was built using LSTM and dense layers to predict the earthquake epicenter as a distance from a predefined location, which provided a root mean square error of 1.25.},
journal = {IEEE Access},
author = {Banna, Md. Hasan Al and Ghosh, Tapotosh and Nahian, Md. Jaber Al and Taher, Kazi Abu and Kaiser, M. Shamim and Mahmud, Mufti and Hossain, Mohammad Shahadat and Andersson, Karl},
year = {2021},
note = {Conference Name: IEEE Access},
keywords = {Attention, Computer architecture, Earthquakes, LSTM, Load modeling, Neural networks, Predictive models, Recurrent neural networks, Support vector machines, earthquake, location, occurrence},
pages = {56589--56603},
}
Downloads: 2
{"_id":"SLQ4HJ9KK3Zf8cQJb","bibbaseid":"banna-ghosh-nahian-taher-kaiser-mahmud-hossain-andersson-attentionbasedbidirectionallongshorttermmemorynetworkforearthquakeprediction-2021","author_short":["Banna, M. H. A.","Ghosh, T.","Nahian, M. J. A.","Taher, K. A.","Kaiser, M. S.","Mahmud, M.","Hossain, M. S.","Andersson, K."],"bibdata":{"bibtype":"article","type":"article","title":"Attention-Based Bi-Directional Long-Short Term Memory Network for Earthquake Prediction","volume":"9","issn":"2169-3536","doi":"10.1109/ACCESS.2021.3071400","abstract":"An earthquake is a tremor felt on the surface of the earth created by the movement of the major pieces of its outer shell. Till now, many attempts have been made to forecast earthquakes, which saw some success, but these attempted models are specific to a region. In this paper, an earthquake occurrence and location prediction model is proposed. After reviewing the literature, long short-term memory (LSTM) is found to be a good option for building the model because of its memory-keeping ability. Using the Keras tuner, the best model was selected from candidate models, which are composed of combinations of various LSTM architectures and dense layers. This selected model used seismic indicators from the earthquake catalog of Bangladesh as features to predict earthquakes of the following month. Attention mechanism was added to the LSTM architecture to improve the model’s earthquake occurrence prediction accuracy, which was 74.67%. Additionally, a regression model was built using LSTM and dense layers to predict the earthquake epicenter as a distance from a predefined location, which provided a root mean square error of 1.25.","journal":"IEEE Access","author":[{"propositions":[],"lastnames":["Banna"],"firstnames":["Md.","Hasan","Al"],"suffixes":[]},{"propositions":[],"lastnames":["Ghosh"],"firstnames":["Tapotosh"],"suffixes":[]},{"propositions":[],"lastnames":["Nahian"],"firstnames":["Md.","Jaber","Al"],"suffixes":[]},{"propositions":[],"lastnames":["Taher"],"firstnames":["Kazi","Abu"],"suffixes":[]},{"propositions":[],"lastnames":["Kaiser"],"firstnames":["M.","Shamim"],"suffixes":[]},{"propositions":[],"lastnames":["Mahmud"],"firstnames":["Mufti"],"suffixes":[]},{"propositions":[],"lastnames":["Hossain"],"firstnames":["Mohammad","Shahadat"],"suffixes":[]},{"propositions":[],"lastnames":["Andersson"],"firstnames":["Karl"],"suffixes":[]}],"year":"2021","note":"Conference Name: IEEE Access","keywords":"Attention, Computer architecture, Earthquakes, LSTM, Load modeling, Neural networks, Predictive models, Recurrent neural networks, Support vector machines, earthquake, location, occurrence","pages":"56589–56603","bibtex":"@article{banna_attention-based_2021,\n\ttitle = {Attention-{Based} {Bi}-{Directional} {Long}-{Short} {Term} {Memory} {Network} for {Earthquake} {Prediction}},\n\tvolume = {9},\n\tissn = {2169-3536},\n\tdoi = {10.1109/ACCESS.2021.3071400},\n\tabstract = {An earthquake is a tremor felt on the surface of the earth created by the movement of the major pieces of its outer shell. Till now, many attempts have been made to forecast earthquakes, which saw some success, but these attempted models are specific to a region. In this paper, an earthquake occurrence and location prediction model is proposed. After reviewing the literature, long short-term memory (LSTM) is found to be a good option for building the model because of its memory-keeping ability. Using the Keras tuner, the best model was selected from candidate models, which are composed of combinations of various LSTM architectures and dense layers. This selected model used seismic indicators from the earthquake catalog of Bangladesh as features to predict earthquakes of the following month. Attention mechanism was added to the LSTM architecture to improve the model’s earthquake occurrence prediction accuracy, which was 74.67\\%. Additionally, a regression model was built using LSTM and dense layers to predict the earthquake epicenter as a distance from a predefined location, which provided a root mean square error of 1.25.},\n\tjournal = {IEEE Access},\n\tauthor = {Banna, Md. Hasan Al and Ghosh, Tapotosh and Nahian, Md. Jaber Al and Taher, Kazi Abu and Kaiser, M. Shamim and Mahmud, Mufti and Hossain, Mohammad Shahadat and Andersson, Karl},\n\tyear = {2021},\n\tnote = {Conference Name: IEEE Access},\n\tkeywords = {Attention, Computer architecture, Earthquakes, LSTM, Load modeling, Neural networks, Predictive models, Recurrent neural networks, Support vector machines, earthquake, location, occurrence},\n\tpages = {56589--56603},\n}\n\n","author_short":["Banna, M. H. A.","Ghosh, T.","Nahian, M. J. A.","Taher, K. A.","Kaiser, M. S.","Mahmud, M.","Hossain, M. S.","Andersson, K."],"key":"banna_attention-based_2021","id":"banna_attention-based_2021","bibbaseid":"banna-ghosh-nahian-taher-kaiser-mahmud-hossain-andersson-attentionbasedbidirectionallongshorttermmemorynetworkforearthquakeprediction-2021","role":"author","urls":{},"keyword":["Attention","Computer architecture","Earthquakes","LSTM","Load modeling","Neural networks","Predictive models","Recurrent neural networks","Support vector machines","earthquake","location","occurrence"],"metadata":{"authorlinks":{}},"downloads":2},"bibtype":"article","biburl":"https://api.zotero.org/users/3831970/collections/RXZ8CCMC/items?key=YLkgSSQIItiwGkufDEgcZ27J&format=bibtex&limit=100","dataSources":["eBFLpkHdRMDh9Juvn","M32K9b97FBA3ABT6f","XxnyXEGfAxbBCrK23","vA5ZqqGu3JauixXfw","7w9HtinE43mdCrTS8","NJ3ihftJtTRdSgxvC"],"keywords":["attention","computer architecture","earthquakes","lstm","load modeling","neural networks","predictive models","recurrent neural networks","support vector machines","earthquake","location","occurrence"],"search_terms":["attention","based","directional","long","short","term","memory","network","earthquake","prediction","banna","ghosh","nahian","taher","kaiser","mahmud","hossain","andersson"],"title":"Attention-Based Bi-Directional Long-Short Term Memory Network for Earthquake Prediction","year":2021,"downloads":2}