Classification and statistical analysis of hydrothermal seafloor rocks measured underwater using laser-induced breakdown spectroscopy

Abstract

This study investigates the use of statistical methods for the classification of laser‐induced breakdown spectroscopy (LIBS) measurements of water‐immersed rocks with respect to their labels and geological groups. The analysis is performed on deep‐sea hydrothermal deposit rocks. These rocks are categorized on the basis of the relative ratio of Zn‐Pb‐Cu on a ternary diagram. The proposed method demonstrates that the accurate classification of rocks with respect to their labels and geological group from the LIBS data can be successfully achieved by combining principal component analysis (PCA), which reduces the dimensionality of the data, with classification algorithms such as the support vector machine (SVM), k‐nearest neighbor search (KNN), and artificial neural network (ANN) methods. The performance of the classification algorithms depends on the size of the dataset. Additionally, removing the linear trend from the data enhances the performance of the classification in terms of sensitivity. The best classification performance concerning the rock label is obtained using an SVM linear kernel algorithm with 95% sensitivity. The best classification performance concerning the geological group is obtained using the SVM method with 98% accuracy. The one‐sided Wilcoxon signed rank test is applied to the classification results in the rock label and group cases, and the results indicate that the SVM algorithm has statistical significance over the other algorithms while classifying the rock labels and rock group.

More information

Identifiers

Catalogue record

Export record

Altmetrics