ACADEMIC RESEARCH IN MODERN SCIENCE
International scientific-online conference
147
RESULTS OF THE INITIAL DATABASE PREPARATION STAGE FOR
DEVELOPING GEOCHEMICAL SEARCH CRITERIA FOR RARE EARTH
ELEMENT MINERALIZATION IN BLACK SHALES BASED ON
ARTIFICIAL INTELLIGENCE
Zafarjon Ergashov
Doctoral student of the Department of Geochemistry and Mineralogy of the
National University of Uzbekistan
e-mail: zafarjonergashov@gmail.com
https://doi.org/10.5281/zenodo.14557687
Nowadays, in all sectors of human activity, in particular finance, economics,
marketing, medicine and engineering the role of information is growing sharply,
and this reason raises the demand for their effective manipulation, processing,
and interpretation to study in various interests.
Today, a number of companies and research institutes are involving several
scientific achievements and advanced technologies in order to improve the
efficiency of work processes and to support successful discoveries. One of these
approaches to make these jobs easier, more powerful, expected results more
natural and less time-consuming to process large amounts of data is applying
Artificial Intelligence (AI).
"Research on processing geochemical data and identifying geochemical
anomalies has made important progress in recent decades. Fractal/multi-fractal
models, compositional data analysis, and machine learning (ML) are three
widely used techniques in the field of geochemical data processing. In recent
years, ML has been applied to model the complex and unknown multivariate
geochemical distribution and extract meaningful elemental associations related
to mineralization or environmental pollution. It is expected that ML will have a
more significant role in geochemical mapping with the development of big data
science and artificial intelligence in the near future..." [1].
Moreover, in accordance with the Strategy "Digital Uzbekistan - 2030" and
to create favorable conditions for the accelerated introduction of artificial
intelligence technologies and their widespread use in the country, the Decree of
the President of Uzbekistan No. PP-4996 on February 17, 2021 came out.
Although there are not so much experience and literature, I strongly believe
that involving new scientific trends such as Data mining (DM) and AI have a
great impact on prospering of all directions of human life, in particular
geoscience.
ACADEMIC RESEARCH IN MODERN SCIENCE
International scientific-online conference
148
Geological information, including the different properties of rocks and the
results for chemical elements are the object of applied geochemistry. And
traditionally, these data are processed and geostatistically analyzed in order to
study geochemical features such as the geochemical specializations of different
rock groups, the determination of ore-generating chemical elements, and their
various geochemical assemblies.
On the basis of these statistical studies, various geochemical maps (mono-
element, additive, multiplicative) are built using methods like Kriging or Inverse
distances to a power, in other words, maps of the halos of chemical elements.
Studying geochemical data that includes a wide range of chemical elements,
it is difficult to determine objectively those geostatistical features listed above
because of their complexity.
"This is because many exploration geologists rely upon the classical
geostatistical method of Kriging which oftentimes do not produce accurate
predictions due to the complexity of interactions between geological features
and spatial variables" [2].
Based on these facts and arguments above, there is a great need to apply
Data mining and Artificial Intelligence methods to geochemical research.
Therefore, I do research on applied aspects of geoscience, and advanced
geological modeling on geochemical data, which helps to discover new
perspective areas of deposits by developing research criteria of ore formations.
In the mineral industry and geological, especially geochemical research, and
great enthusiasm for studying geological data, I attempt to pursue my
dissertation on the topic: “Development of geochemical search criteria for rare
earth element mineralization in black shales based on artificial intelligence"
within my doctoral program.
To achieve the designated goal of developing geochemical search criteria,
ICP-MS
(61 elements) analysis results from over 7,000 samples taken from various
mining workings (lithogeochemical sampling along primary halos, trenches, core
drill holes, and adits) were used.
Initially, the Ustuk area database was prepared in MS Excel for processing
in computer programs. Geostatistical analysis was mainly performed using
Statistica 10 and QGIS, while geospatial analysis and geochemical mapping work
was carried out using ArcGIS, QGIS, Global Mapper, and primarily Surfer 18
software.
ACADEMIC RESEARCH IN MODERN SCIENCE
International scientific-online conference
149
In the initial stage, mainly using MS Excel, all chemical element laboratory
analysis results (ICP-MS, 61 elements) were converted to the same unit of
measurement (g/t). For results below laboratory analysis sensitivity or
extremely high quantities, the lower and upper percentile (10%) values were
mainly adopted. Samples with missing analysis results were removed from the
database. As a result, from the initial 8,496 samples, 7,256 were retained for use
in subsequent stages. In the next stage, coordinates were calculated for each
sample from mining workings (lithogeochemical sampling along primary halos,
trenches, core drill holes, and adits) needed for geospatial analysis. Samples in
the database were categorized based on rock descriptions and rock-forming
chemical elements.
Methodology and expected results:
Traditional and new approaches based on DM and AI of geostatic analysis
will be performed and the necessary graphics (diagrams, halo maps, block
diagrams) will be built.
In the next stages of the study, the results of both approaches (geostatistics
parameters, geochemical patterns, geochemical halos, etc.) and the time spent
and quality of the results are compared, and work on drawing conclusions is
ongoing.
References:
1.
Sklyarov E.V. "Interpretation of Geochemical Data" (Moscow, 2001);
2.
Grigoryan S.V. "Primary Geochemical Halos in Prospecting and Exploration
of Ore Deposits" (Moscow, 1987);
3.
Renguang Zuo. Machine Learning of Mineralization-Related Geochemical
Anomalies: A Review of Potential Methods, May 2017, Natural Resources
Research 26(4):1-8;
4.
Fareed Majeed, Yao Yevenyo Ziggah, Charles Kusi-Manu, Bemah Ibrahim,
Isaac Ahenkorah. A novel artificial intelligence approach for regolith
geochemical grade prediction using multivariate adaptive regression splines,
February 2022, Geosystems and Geoenvironment 1(4):100038.
