Artificial intelligence and machine learning

Research Article

Machine learning-based mineralogical analysis using planimetric method

Maksim Gennad'evich Shishaev¹, Vladimir Vital'evich Dikovitsky²

1,2	Putilov Institute for Informatics and Mathematical Modeling of the Kola Science Centre of the Russian Academy of Sciences, Apatity, Russia
1	m.shishaev@ksc.ru

Abstract. The problem of automation of mineralogical analysis by the planimetric method is considered using the analysis of various types of apatite ores typical of the Khibiny deposit as an example.

Purpose: To study the efficiency of machine learning as a means for forming a features vector and solving the classification problem in various formulations during planimetric analysis of minerals.

Results: The study revealed the features of planimetric analysis as a classification problem, the properties of adjacency and homogeneity of classes in the space of identifying features are determined. Potential systematic errors in determining the content of the valuable component by the planimetric method are analyzed for various formulations of the classification problems. Experiments confirmed the possibility of directly using the pre-trained convolutional network ResNet-18 for forming a feature vector of classification objects, ensuring good separability of classes. Using the example of the ores under consideration, the high efficiency (over 98% precision) of the neural network classifier and vectorizer ResNet-18 for identifying image elements related to the pure classes "apatite" / "non-apatite" is experimentally confirmed. High classification precision is maintained with planimetric grid cell sizes down to 2$\times$ 2 pixels (78%), and approaches 100% at 20$\times$ 20 pixels.

The effectiveness of a neural network approach to determining the specific grade of a valuable component in ore was studied. Experiments did not confirm the effectiveness of implementing planimetric analysis as a soft classification problem without significant modifications to the classifier’s architecture. However, they demonstrated the high effectiveness of the approach in a multi-class setting of the problem. The absolute error in determining the grade of a valuable component in the latter case depends on the number of classes and the ore type and, in the worst case, does not exceed 6%, which is higher than the accuracy of expert assessments by experienced mine geologists.

Practical relevance: the approach is applicable to the development of inexpensive, fast, and efficient express ore analyzers that do not require specialized equipment. (In Russian).

Keywords: planimetric mineralogical analysis, machine learning, ResNet18, classification

MSC-2020 68T45; 00A99

Acknowledgments: The study was carried out within the framework of the Putilov Institute for Informatics and Mathematical Modeling of the Kola Science Centre of the Russian Academy of Sciences state assignment of the Ministry of Science and Higher Education of the Russian Federation, research topic: "Methods and technologies for creating intelligent information systems to support the development of complex dynamic systems with regional specificity in conditions of uncertainty and risk" (topic code FMEZ-2025-0053)

For citation: Maksim G. Shishaev, Vladimir V. Dikovitsky. Machine learning-based mineralogical analysis using planimetric method. Program Systems: Theory and Applications, 2025, 16:4, pp. 241–266. (In Russ.). https://psta.psiras.ru/2025/4_241-266.

Full text of article (PDF): https://psta.psiras.ru/read/psta2025_4_241-266.pdf.

The article was submitted 12.10.2025; approved after reviewing 13.10.2025; accepted for publication 29.10.2025; published online 14.11.2025.