Evgenii Maltcev


Chief Specialist
Institute TOMS

The accumulated 20 years of experience and special skills allow us to solve the problems of mining and geological modeling and evaluation of mineral deposits (according to Russian and international standards), as well as to assess the representativeness, quality and reliability of geological exploration data (QA/QC). Also (in the pre-computer era) experience of several years of work as a district geologist in the exploration and production of the Bamskoye gold deposit (Amur region). Recently, several projects have been carried out to introduce neural network technologies and machine learning methods into the process of creating a geological and technological block model and developing geological and technological mapping programs based on high expressiveness and reliability of estimates of technological indicators of ores combined with lower labor intensity of work. Author of more than 10 articles. NAEN expert.

Application of machine learning algorithms for solving tasks of geo-metallurgical modeling and interwell interpolation

Due to the heterogeneity of the ore composition, the technological parameters, as well as the content of the useful component, are a variable value distributed in a certain way in three-dimensional space.
The block model provides complete information about spatial heterogeneity.
A significant reduction in time and production costs with a high degree of accuracy and reliability of the forecast of geological and technological indicators, as well as the maximum effect due to the use of spatial three-dimensional models is the ultimate goal and result of modern modeling methods.
Analysis of the advantages of machine learning methods for interpolation and forecasting problems in the inter-well space.