Machine Learning Interpretation of Conventional Well Logs in Crystalline Rocks

摘要: The identification of lithologies is a crucial task in continental scientific drilling research. In fact, complex geological situations such as crystalline rocks, more nonlinear functional behaviors exist well log interpretation/classification purposes; thus posing challenges accurate lithology using geophysical data the context rocks. aim this work to explore capability k-nearest neighbors classifier and demonstrate its performance comparison with other classifiers results show that best was neural network followed by support vector machine neighbors. These intelligence learning methods appear be promising recognizing can very useful tool facilitate geophysicists allowing them quickly get nature all units during exploration phase.