Autoregressive Conditional Duration Modeling and Forecasting the Earthquakes Temporal Data of the Different Regions of Pakistan
ACTA GEOPHYSICA(2024)
Center for Earthquake Studies | Ludwig Maximilian University
Abstract
Pakistan is situated in an earthquake-prone area, and in the past, many destructive earthquakes have occurred including the most destructive 1935 Quetta, and 2005 Kashmir, earthquakes. Therefore, it is of vital importance to investigate the statistical properties of the earthquake temporal data of Pakistan. The present study brings novelty through the vast extension of the linear and nonlinear autoregressive conditional duration models in the field of seismology. A simple duration model capable of describing and forecasting the earthquake elapsed times data has been developed. Different specifications of the duration models were considered to completely extract the time-dependent structure in earthquake data. This study aims to identify the most suitable in-sample fitted and out-of-sample forecasting models for the earthquake elapsed times data of Pakistan. A variety of autoregressive conditional duration models were applied to the complete and updated earthquake catalog of Pakistan. The most suitable model was chosen through statistical model evaluation techniques. The method of maximum likelihood was used to estimate the model parameters. The adequacy of fitted models is assessed through residual analysis. The obtained results suggest that the Logarithmic Autoregressive Conditional Duration model of type 2 (LACD2) appeared as the most suitable in-sample fitted model for describing the earthquake temporal data of the different zones of Pakistan. Further, the simple autoregressive conditional duration (ACD) model outperforms the remaining considered models regarding 1-through-4 steps-ahead out-of-sample forecasting performance for both India–Eurasia collision (IEC) and Makran subduction (MS) zones. Autoregressive models with Burr and exponential distributions as assumed error distributions have appeared as the most suitable fitted models for IEC and MS zones, respectively. The residuals analysis results show that the most suitable fitted models are correctly identified. The results show that the earthquake short-term (h=1) forecasting with duration models is more accurate in comparison with earthquake long-term (h=1, 2, 3) forecasting. The forecasted elapsed times for IEC and MS zones are 0.60 and 2.94 years, respectively. The obtained results show that the autoregressive conditional models are a more useful tool for forecasting the earthquake elapsed times for short term in comparison with the long-term forecasting. Hence, autoregressive conditional models are capable of modeling and forecasting the earthquake temporal data of the different regions of Pakistan. Among the competing models, the best fit model can serve the purpose of data description, missing value estimation in earthquake catalogs and uncertainty quantification in the earthquake occurrence process. The most suitable forecasted model yields the future earthquake occurrence trend in the study region.
MoreTranslated text
Key words
Temporal data,Autoregressive conditional duration models,Autocorrelation,Model selection,Diagnostic analysis,Forecasting
求助PDF
上传PDF
View via Publisher
AI Read Science
AI Summary
AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.
Example
Background
Key content
Introduction
Methods
Results
Related work
Fund
Key content
- Pretraining has recently greatly promoted the development of natural language processing (NLP)
- We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
- We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
- The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
- Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance
Upload PDF to Generate Summary
Must-Reading Tree
Example

Generate MRT to find the research sequence of this paper
Related Papers
1998
被引用1196 | 浏览
2007
被引用236 | 浏览
1984
被引用177 | 浏览
1988
被引用2742 | 浏览
2000
被引用388 | 浏览
2005
被引用15 | 浏览
2009
被引用84 | 浏览
1974
被引用1279 | 浏览
2015
被引用11 | 浏览
2018
被引用18 | 浏览
2017
被引用8 | 浏览
2017
被引用18 | 浏览
2020
被引用3 | 浏览
Data Disclaimer
The page data are from open Internet sources, cooperative publishers and automatic analysis results through AI technology. We do not make any commitments and guarantees for the validity, accuracy, correctness, reliability, completeness and timeliness of the page data. If you have any questions, please contact us by email: report@aminer.cn
Chat Paper
GPU is busy, summary generation fails
Rerequest