Reasoning Geo-Spatial Neutral Similarity from Seismic Data Using Mixture and State Clustering Models
Conventionally, earthquake events are recognized by guided and well established geographical region confines. However, explicit regional schemes are prone to overlook patterns manifested by cross-boundary seismic relations that are regarded vital to seismological research. Rather, we investigate a statistically motivated system that clusters earthquake impacted places by similarity in seismic feature space, and is hence impartial to geo-spatial proximity constraints. To facilitate our study, we have acquired hundreds of thousands recordings of earthquake episodes that traverse an extended time period of forty years. Episodes are split into groups singled out by their affiliated geographical place, and from each, we have extracted objective seismic features expressed in both a compact term-frequency of scales format, and as a discrete signal representation that captures magnitude samples spaced in regular time intervals. Attribute vectors of the distributional and temporal domains are further applied towards our mixture model and Markov chain frameworks, respectively, to conduct clustering of presumed unlabeled, shake affected locations. We performed comprehensive cluster analysis and classification experiments, and report robust results that support the intuition of geo-spatial neutral similarity.
KeywordsEarthquake Seismic Mixture model Hidden Markov model Expectation-maximization Clustering k-nearest neighbors
We would like to thank the anonymous reviewers for their insightful and helpful feedback on our work.
- 1.Akaike, H.: Information theory and an extension of the maximum likelihood principle. In: International Symposium on Information Theory, Budapest, pp. 267-281 (1973)Google Scholar
- 2.Baeza-Yates, R., Ribeiro-Neto, B.: Modern Information Retrieval. ACM Press Series/Addison Wesley, Essex (1999)Google Scholar
- 3.Baum, L.E.: An inequality and associated maximization technique in statistical estimation for probabilistic functions of Markov processes. In: Symposium on Inequalities, Los Angeles, pp. 1-8 (1972)Google Scholar
- 7.Duda, R.O., Hart, P.E., Stork, D.G.: Unsupervised learning and clustering. In: Pattern Classification, pp. 517–601. Wiley, New York (2001)Google Scholar
- 8.Flinn-Engdahl Seismic and Geographic Regionalization Scheme (2000). http://earthquake.usgs.gov/learn/topics/flinn_engdahl.php
- 11.GeoJSON Format for Encoding Geographic Data Structures (2007). http://geojson.org/
- 17.Manning, C.D., Schutze, H.: Foundations of Statistical Natural Language Processing. MIT Press, Cambridge (2000)Google Scholar
- 21.R Project for Statistical Computing (1997). http://www.r-project.org/
- 27.United States Geological Survey (USGS) (2004). http://earthquake.usgs.gov/earthquakes/feed/v1.0/