Perception And Administering Regression Analysis On Earthquake Data For Finding An Appropriate Correlation: Case Study On Nepal Earthquake

Abstract

A popular statistical method, Regression Analysis is built between a dependent
variable and more than one independent variable/s to generate the relationship and correlation
which can be used for varied predictions. For estimating the same, regression analysis is
applied statistically, mathematically and justifies to conclude a logical result calculating
Residual Value (RV) and R Square in regression for earthquake prediction. The primary focus
of this paper is to produce better prediction on earthquake in terms of resilience, versatility,
accuracy, authenticity so that emphasis is applied on nonlinear, linear and multilinear
regression model having free hand curve, least square method, standard deviation (SD) using
arithmetic mean and assumed mean and statistical analysis. Keeping this result in mind,
multilinear regression model is achieved by using historical data of Nepal statistically for five
years with parameters like the magnitude, the location, the date, the time, the depth etc. and
established a relationship like the dependency of depth and magnitude on earth’s crust and
temperature on probability of occurrences of earthquake through ANOVA table. This study
also recommends that that seismic waves of the earth’s crust dependency on three of the parameters are analyzed are very significant predictors for the occurrences of earthquake. We
could also find significant correlations between the analyzed indicators..