Page 322 - Machine Learning for Subsurface Characterization

Representative materials containing discontinuities of various spatial distributions generated based on random, linear, unimodal Gaussian, and bimodal
are char- noninvasive the for classifiers train to datasets four the generate to used Gaussian. double (D) and Gaussian, single (C) linear, (B) Random, (A)
plots distributions.
upper
in various
presented of

those

to discontinuities
similar static

Materials containing

functions. materials

intensity of

FIG. 9.25 Gaussian acterization