CWT was applied for lithology detection over the Costa Rica margin, using gamma ray logs as a best indicator. It is observed that four major sedimentary layers terminated there with a gabbroic sill. DWT was applied on density logs to identify small changes in the rock matrix. Accordingly, wavelet-based fractal analysis was applied on every sedimentary layer. This analysis indicates that a certain portion of wells contain abundant shale and low-energy environments.

Successful clustering of a dataset which is consistent with geological facts and laboratory and field studies is one of the results of this study. Since only REEs were used in this division, a good agreement of the results with lithology is considerable. Results show that unsupervised pattern recognition helps find some hidden information which would be difficult to obtain in usual ways. In addition, methods presented in this study will enable better interpretation of data.

Particle swarm optimization is developed to estimate the model parameters of a 2.5-D sedimentary basin. PSO have been implemented on synthetic data and two field data. An observation has been made that PSO is affected by some levels of noise, but estimated depths are close to the true depths. The PSO results are well correlated with borehole samples and provide more geological viability than Marquardt results. Despite its long computation time, it is very simple to implement.