Comparative study of spatial autocorrelation methods in the statistical analysis of deposit mineralization patterns

A comparison of spatial autocorrelation methods used to form perspective classes within the heterogeneous spatial distribution of mineralization of the deposit is carried out. Classifiers, which are key in mapping the prospects of minerals based on data, allow us to develop models for predicting the mineral potential of poorly studied sites within known mining areas. The paper presents a comparative study of methods of statistical processing of spatial data using the Moran's index, Getis-Ord index, and the author’s approach based on a modified Moran's index. The methods have been tested on the basis of geochemical data from the Agbau gold mining complex in Côte d'Ivoire, provided by the Endeavour Mining Corporation campaign. An assessment of the reliability of the results obtained with reference data has been carried out, taking into account the current operating conditions of the field. The results were obtained in terms of performance indicators: the author's approach: accuracy = 98%, F1-score = 98% (at 250 pbb) and the stability of the F1-score (from 93 to 96%) when changing the onboard content; the approach based on the classical Moran index: accuracy = 77%, F1-score = 85% (at 250 pbb) and a decrease in F1-score (from 93 to 66%); approach based on the Goethis-Ord index: accuracy = 75%, F1-score = 84% (at 250 pbb) and a decrease in F1-score (from 90 to 55%). The author's approach It demonstrated resistance to fluctuations in onboard content, unlike classical methods.

Nkrumah A. H. M., Silina T. S., Fayzrakhmanov R. A. Comparative study of spatial autocorrelation methods in the statistical analysis of deposit mineralization patterns. MIAB.MiningInf.Anal.Bull.2025;(12-1):170-185. [InRuss]. DOI: 10.25018/0236_1493_2025_ 121_0_170.

Assoumou Herve Mathieu Nkrumah¹, Graduate Student, e-mail: nkvetcho@gmail.com,
T.S. Silina, Cand. Sci. (Geol. Mineral.), Assistant Professor, Ural State Mining University, 620144, Ekaterinburg, Russia, e-mail: tamarasil @mail.ru,
R.A. Faizrakhmanov¹, Dr. Sci. (Econ.), Professor, Head of Chair, e-mail: Fayzrakhmanov@gmail.com,
¹ Perm National Research Polytechnic University, 614990, Perm, Russia.

A.H.M. Nkrumah, e-mail: nkvetcho@gmail.com.

1. Stöferle R. P., Valek M. J. Stagflation 2.0: In gold we trust report. Liechtenstein, Incrementum, 2022, XVI, 393 p. 

2. Nekrasov Е. М. Principles governing the selection of ore regions for gold deposit search. Pro¬ceedings of higher educational establishments. Geology and exploration. 2022, no. 63(6), pp. 77—86. [In Russ]. DOI: 10.32454/0016-7762-2020-63-6-77-86. 

3. Piskorsky N. P. Mineralogical and geochemical models of zoning and local forecast of gold min¬eralization. Gold mining. 2022, no. 11(288). [In Russ]. https://zolotodb.ru/article/12883. 

4. Rustamov A. R., Penkov G. M., Petrakov D. G., Rustamova M. A. Modern methods of using machine learning as a tool for oil production forecasting. Perm Journal of Petroleum and Mining Engi¬neering. 2024, vol. 24, no. 1, pp. 44—50. [In Russ]. DOI: 10.15593/2712-8008/2024.1.6. 

5. Rezapoor M., Abedi M., Bahroudi A., Rahimi H. A clustering approach for mineral potential map¬ping: A deposit-scale porphyry copper exploration targeting. Geopersia. 2020, vol. 10, no. 1, pp. 149— 163. DOI: 10.22059/GEOPE.2019.284414.648488. 

6. Daviran M., Maghsoudi A., Cohen D., Ghezelbash R., Yilmaz H. Assessment of various fuzzy c-mean clustering validation indices for mapping mineral prospectivity: Combination of multifractal geochemical model and mineralization processes. Natural Resources Research. 2019, vol. 29, pp. 229— 246. DOI: 10.1007/s11053-019-09571-9. 

7. Carranza E. J. M., Laborte A. G. Random forest predictive modeling of mineral prospectivity with small number of prospects and data with missing values in Abra (Philippines). Computers & Geo¬sciences. 2015, vol. 74, pp. 60—70. DOI: 10.1016/j.cageo.2014.10.004. 

8. Ram A., Sunita J., Jalal A. S., Kumar M. A density based algorithm for discovering density varied clusters in large spatial databases. International Journal of Computer Applications. 2010, vol. 3, no. 6. DOI: 10.5120/739-1038. 

9. Krasnykh S. S. Spatial autocorrelation model for the implementation of digital technologies in the business of the russian federation. Bulletin of the Academy of Knowledge. 2020, no. 4 (39), pp. 228—235. [In Russ]. DOI: 10.24411/2304-6139-2020-10467. 

10. Khamidullin R. A., Brekhova I. N. Characteristics of types and kinds of geological maps. Actual problems of humanitarian and natural sciences. 2017, no. 2(4), pp. 72—74. [In Russ]. 

11. Bragin V. I., Kharitonova M. Yu., Matsko N. A. A probabilistic approach to the assessment of dynamic onboard contents. Journal of Mining Institute. 2021, vol. 251, pp. 617—625. [In Russ]. DOI: 10.31897/PMI.2021.5.1. 

12. Rezapoor M., Abedi M., Bahroudi A., Rahimi H. A clustering approach for mineral potential mapping: A deposit-scale porphyry copper exploration targeting. Geopersia. 2019, vol. 10, pp. 146— 163. DOI: 10.22059/GEOPE.2019.284414.648488. 

13. Mitin G. V., Panov A. V. Clustering of one-dimensional streaming data based on the density of data distribution in the feature space. Electronic scientific journal «IT-Standard». 2024, no. 1, pp. 18—33. [In Russ]. 

14. Shaw K. O., Goïta K., Germain M. Prospectivity mapping of heavy mineral ore deposits based upon machine-learning algorithms: Columbite-tantalite deposits in west-central côte d’ivoire. Minerals. 2022, vol. 12, no. 11, article 1453. DOI: 10.3390/min12111453. 

15. Williams M. V., Kuzyakov O. N. Application of Getis-Ord Gi* and DBSCAN method when prospecting for mineral resources. Automation and informatization of the fuel and energy complex. 2025, no. 1(618), pp. 35—41. [In Russ]. 

16. Hronsky J. M. A., Kreuzer O. P. Applying spatial prospectivity mapping to exploration target¬ing: Fundamental practical issues and suggested solutions for the future. Ore Geology Reviews. 2019, vol. 107, pp. 647—653. DOI: 10.1016/j.oregeorev.2019.03.016. 

17. Kashirtseva A. P. Calculation of cut-off grade in the preparation of feasibility studies for the development of gold ore deposits. Mathematical and Computer Modeling in Economics. 2016, no. 1, pp. 233—239. [In Russ]. 

18. Nkrumah A. H. M., Fayzrakhmanov R. A., Silina T. S. Methodology and algorithm for analyzing the patterns of spatial distribution of mineralization. dynamic spatial autocorrelation approach. Perm National Research Polytechnic University Bulletin. Electrotechnics, information technologies, control systems. 2025, no. 2, pp. 150—160. [In Russ]. 

19. Fomin S. I., Govorov A. S. Validation of the chosen cutoff grade value in open pit mine design. MIAB. Mining Inf. Anal. Bull. 2023, no. 12, pp. 169—182. [In Russ]. DOI: 10.25018/0236_1493_ 2023 12 0 169. 

20. Anselin L. Local indicators of spatial association—LISA. Geographic Analysis. 1995, vol. 27, pp. 93—115. DOI: 10.1111/j.1538-4632.1995. tb 00338.x. 

21. Kupriyanov A. O. Coordinate transformations in the design of extended objects. Perspectives of Science and Education. 2016, no. 1(19), pp. 53—57. [In Russ]. 

22. Timofeev M. V. Method of testing a hypothesis in applied marketing problems using an error matrix. Stolypinsky Vestnik. 2022, no. 9, pp. 4830—4840. [In Russ].