An unsupervised learning algorithm: application to the discrimination of seismic events and quarry blasts in the vicinity of Istanbul

Abstract

Abstract.The results of the application of an unsupervised learning (neural network) approach comprising a Self Organizing Map (SOM), to distinguish micro-earthquakes from quarry blasts in the vicinity of Istanbul, Turkey, are presented and discussed. The SOM is constructed as a neural classifier and complementary reliability estimator to distinguish seismic events, and was employed for varying map sizes. Input parameters consisting of frequency and time domain data (complexity, spectral ratio, S/P wave amplitude peak ratio and origin time of events) extracted from the vertical components of digital seismograms were estimated as discriminants for 179 (1.8 < M d < 3.0) local events. The results show that complexity and amplitude peak ratio parameters of the observed velocity seismogram may suffice for a reliable discrimination, while origin time and spectral ratio were found to be fuzzy and misleading classifiers for this problem. The SOM discussed here achieved a discrimination reliability that could be employed routinely in observatory practice; however, about 6% of all events were classified as ambiguous cases. This approach was developed independently for this particular classification, but it could be applied to different earthquake regions.

Application of &lt;i&gt;k&lt;/i&gt;-means and Gaussian mixture model for classification of seismic activities in Istanbul

“…Thus, we compare their responses with pre-defined targets that were selected by authors manually. Kuyuk et al (2011) applied an unsupervised algorithm, called selforganizing map (SOM) as a neural classifier for the same region using the partially similar discriminants. Although they used extra two parameters (spectral ratio and origin time of events) for better classification, their results indicated that these two are fuzzy and misleading classifiers.…”

Detecting Seismic Anomalies in Outgoing Long-Wave Radiation Data

“…Meanwhile, much previous work has been done on surface latent heat flux (SLHF) precursors [22], electromagnetic signal precursors [17], [18], [23], [24] and electric earthquake precursors [25]. Kuyuk et al used unsupervised learning (neural networks) to obtain a self-organizing map in order to distinguish micro-earthquakes from quarry blasts [26]. Giacco et al proposed a classification method based on the multilayer perceptron and SVM, and used the proposed method to classify seismic signals into three different seismic events: 1) explosion-quake; 2) landslide; and 3) volcanic micro-tremor signals [27].…”

Systematic identification of external influences in multi-year microseismic recordings using convolutional neural networks

“…Passive monitoring of elastic waves, generated by the rapid release of energy within a material (Hardy, 2003) is a nondestructive analysis technique allowing a wide range of applications in material sciences (Labuz et al, 2001), engineering (Grosse, 2008) and natural hazard mitigation (Michlmayr et al, 2012) with recently increasing interest into investigations of various processes in rock slopes (Amitrano et al, 2010;Occhiena et al, 2012). Passive monitoring techniques may be broadly divided into three categories, characterized by the number of stations (single vs. array), the duration of recording (snapshot vs. monitoring) and the type of analysis (continuous vs. event-based).…”

“…Due to its simplicity, this event detector is commonly used to assess seismic activity by calculating the number of triggering events per time interval for a time period of interest (Withers et al, 1998;Amitrano et al, 2005;Senfaute et al, 2009). It is often used in the analysis of unstable slopes (Colombero et al, 2018;Levy et al, 2011) and is available integrated into many commercially available digitizers and data loggers (Geometrics, 2018). With respect to unwanted signal components, STA/LTA has also been used to detect external influence factors such as footsteps (Anchal et al, 2018) but due to its inherent simplicity, it cannot reliably discriminate geophysical seismic activity from external (unwanted) influence factors such as noise from humans and natural sources like wind, rain or hail without manually supervising and intervening in the detection process on a case-bycase basis.…”