![]() Detailed analysis showed, that magnitude of completeness for regional scale is MLc=2.5, and for local scale (for example – volcano seismic monitoring) – MLc=–0.2.Ībstract: Global warning shows unpredictablenature that changes in subsurface geologicfeatures. The complex studies for seven the strongest ones were conducted. Kamchatka Seismic Monitoring and Earthquake Prediction System in 2016-2020 allowed registering and processing over 83 thousand tectonic and volcanic earthquakes. The service of automatic data exchange with external users was created and incorporated in SDIS. Creation and development of the Seismological Data Informational System (SDIS) provide the access to seismic observations re-sults for research community. Development algorithms and software for data processing and seismic regime controlling was continued. In particular, the system of data storage was deeply modernized, high-speed access to the data archive was provides, high-performance computing clusters were deployed, all seismic stations were combined in the unified network. One of main results is creation basic informational space, that includes all pro-cesses if seismic observations, from data acquiring till exchange (including external users) of da-ta processing results. ![]() The main direction in the System evolution in this period concerned the creation and modernization of data acquiring and pro-cessing methods. In addition, the retrospective of development of hardware, equipment and software of the System performed. In this paper we present brief review of results of Kamchatka Seismic Monitoring and Earthquake Prediction System operations in the last five years. The experiment results show that the quality of earthquake prediction using SZ-SMOTE as oversampling algorithm significantly outperforms that of using the other oversampling algorithms. The performance of SZ-SMOTE is compared against no oversampling, SMOTE and its popular modifications adaptive synthetic sampling (ADASYN) and borderline SMOTE (B-SMOTE) on six different classifiers. SZ-SMOTE generates synthetic samples with a concentration mechanism in the hyper-sphere area around each selected minority instances. In this paper, we propose a Safe Zone Synthetic Minority Oversampling Technique (SZ-SMOTE) oversampling method as an enhancement of the SMOTE data generation mechanism. But the most popular oversampling methods generate synthetic samples along line segments that join minority class instances, which is not suitable for earthquake precursor data. The general method is to generate more artificial data for the minority class that is the earthquake occurrence data. Since even if an area is in an earthquake-prone zone, the proportion of days with earthquakes per year is still a minority. Addressing Imbalanced Data: SMOTE vs.Earthquake prediction based on extreme imbalanced precursor data is a challenging task for standard algorithms.SMOTE: Synthetic Minority Over-sampling Technique.In this example, we create an imbalanced dataset, split it into training and testing sets, and apply SMOTE to the training data to balance the class distribution. fit_resample ( X_train, y_train ) print ( "Resampled dataset class distribution:", Counter ( y_train_resampled )) # Install the imbalanced-learn library ! pip install - U imbalanced - learn import numpy as np from sklearn.datasets import make_classification from sklearn.model_selection import train_test_split from imblearn.over_sampling import SMOTE from collections import Counter # Create an imbalanced dataset X, y = make_classification ( n_classes = 2, class_sep = 2, weights =, n_features = 20, n_samples = 1000, random_state = 42 ) print ( "Original dataset class distribution:", Counter ( y )) # Split the dataset into training and testing sets X_train, X_test, y_train, y_test = train_test_split ( X, y, test_size = 0.5, random_state = 42 ) # Apply SMOTE to the training data sm = SMOTE ( random_state = 42 ) X_train_resampled, y_train_resampled = sm.
0 Comments
Leave a Reply. |