Page 88 - Contributed Paper Session (CPS) - Volume 1

CPS1110 Selamawit M. et al.
likelihood estimation criterion. The sensitivity of the kernel likelihood model are
tested out on the prediction of posterior model based on the traditional and trend
prior models.
The conclusion is that the trend prior model and the circular uniform kernel
likelihood model provide the best prediction results. Using a reliable kernel
likelihood model appears as the most important feature. Inversion based on
Gaussian kernel likelihood models provide the by far less reliable predictions. Lastly,
in order to have sufficiently good prediction of the sub-surface profiles, one may
need to draw attentions to the choice of band width.

References
1. Baum. L.E., Petrie. T., Soules. G., Weiss. N. (1970). A maximization technique
occurring in the statistical analysis of probabilistic functions of Markov chains.
The Annals of Mathematical Statistics 41(1):164-171
2. Eidsvik. J., Mukerji. T., Switzer. P. (2004). Estimation of geological
attributes from a well log: An application of hidden Markov chains.
Mathematical Geology 36(3):379-396
3. Izenman. A. (1991). Review papers: Recent developments in non
parametric density estimation. Journal of the American Statistical
Association 86(413):205-224.
4. Krumbein. Y.C., Dacey. M.F. (1969). Markov chains and embedded Markov
chains in geology. Mathematical Geology 1(1):79-96 Journal of Petroleum
Science and Engineering 134:237-246
5. Moja. S.S., Asfaw. Z.G., Omre. H. (2018). Math Geosci.
https://doi.org/10.1007/ s11004-018-9752-z
6. Robertson. P.K. (2010). Soil behavior type from the CPT: an update.
California, USA
7. Viterbi. A. (1967). Error bounds for convolutional codes and an
asymptotically optimum decoding algorithm. IEEE transactions on

Information Theory 13(2):260-269

77 | I S I W S C 2 0 1 9