Comparative Study SVM and Random Forest Algorithms for the Classification of Terrestrial Visual Rock Types

Abstract

Abstract The advancement in image processing technology, along with the increasing demand for classifying terrestrial visual rock types in diverse applications such as geological exploration and natural resource mapping, has prompted investigations into the application of machine learning algorithms to enhance classification accuracy. This research aims to assess the performance of the Support Vector Machine (SVM) and Random Forest algorithms in classifying terrestrial visual rock types. The study utilizes a dataset containing images of various rock types sourced from KAGGLE. Both Random Forest and Support Vector Machine methods are employed for rock type classification, and their performances are compared to determine the more effective method. Through experimental analysis on the KAGGLE rock dataset, the study provides evidence supporting the efficacy of the proposed method and identifies the Random Forest algorithm as the more suitable option for rock type classification. Evaluation criteria include accuracy, precision, recall, and F1-score metrics. The research findings reveal that the Random Forest algorithm achieves a higher accuracy rate of 86.25%, with a precision of 0.97, recall of 0.86, and an F1-Score of 0.91, establishing its superiority in rock type classification.