Exploration Geophysics, Machine Learning, and 3D Modeling: Unveiling My Doctoral Thesis!

Full Title of the Ph.D. Thesis: Integrated Imaging through 3D Geophysical Inversion, Multivariate Feature Extraction and Spectral Feature Selection

Summary:

When it comes to mineral exploration geophysics, there are various technical challenges related to data gathering, inverse modeling, model fusion, and integrated interpretations. This study aims to address these challenges by testing and evaluating new methodologies for 3D integrated interpretation of multiple geophysical images with the goal of locating mineral deposits. To achieve this, the study uses 3D inverse modeling of multiple geophysical data, multivariate statistics, and machine learning methods to develop a robust 3D integrated interpretation methodology. The study proposes a cooperative inversion of multiple geophysical data to create a 3D image of an epithermal Au-Ag deposit in British Columbia (Canada). The preliminary integrated interpretation identified four petrophysical domains based on the three cooperatively inverted physical properties: electrical resistivity, IP chargeability, and magnetic susceptibility. Additionally, the study developed a 3D statistical tool to extract geological information from inverted physical property models based on independent component analysis through negentropy maximization. With this methodology, the interpretation process is automated, making it more efficient and accurate.