Year: 2024
Type: conference-paper
Abstract: This work is devoted to the classification of rock types based on their geochemical characteristics using machine learning methods. The study used data on the content of various elements in rocks to develop classification models. Four methods were investigated and compared: decision tree, logistic regression, random forest and gradient boosting. The results showed that the random forest model demonstrates the highest classification accuracy (0.832612), which is explained by its ability to efficiently process a variety of features and their interactions. Correlation analysis has shown significant correlations between the geochemical characteristics of rocks, which underlines the importance of choosing appropriate machine learning methods for processing such data. This work highlights the importance of using ensemble methods that can take into account complex interactions between features for accurate classification of geochemical data and can be useful for specialists in the field of geology, mining and related industries.
Source: E3S Web of Conferences
Language: English
Subfield: Geochemistry and Petrology
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4403775701