Figure 3. The workflow for applying machine learning algorithms.

Dataset acquisition, data analysis and output. The chosen dataset, sourced from descriptions literature on Acantholaimus and Sabatieria species, was organized into matrix labels representing individuals and their corresponding morphological and morphometric characteristics. This organized data served as the input for the subsequent machine-learning stages. The selection and classification algorithms employed encompassed Random Forest, Stochastic Gradient Boosting, Support Vector Machine, and K-nearest neighbor techniques. These algorithms were utilized to identify the optimal set of features for species recognition and to construct predictive models for accurately identifying individuals based on the presence/absence of morphological and morphometric characteristics.