Russian scientists have developed an artificial neural network for mineral research
Release date: 2017-09-12 The Russian Federation’s Research Institute of Scientific Research has released a report that researchers at the Kola Science Center of the Russian Academy of Sciences have used artificial neural networks to map the distribution of complex deposits. This research was successfully published in the famous international journal Science Report. Source: Ministry of Science and Technology Tetanus Toxoid Vaccine,Toxoid Vaccine,Hep B Immune Globulin,Immunoglobulin Injections FOSHAN PHARMA CO., LTD. , https://www.full-pharma.com
In geological mapping, it is more or less affected by objective factors and subjective consciousness. Geologists select rocks according to a certain network (for example, every 50 meters) and break them down into small categories according to their own beauty. This is the primary source of drawing subjectivity, because any set of facts can be Different ways to classify. When the samples are sorted (placed in small piles), geologists begin to determine how the boundaries between these various types of rocks in space are distributed to map and profile the deposits. This is due to my own understanding of the mineral source, comparison with other similar mines, and my own impression of the deposit. Generally speaking, it is related to the degree of education, work experience, a certain school, and aesthetics. As a result, in practice, different geologists of the same deposit can be depicted in completely different ways.
In order to overcome the subjectivity of the mapping, the geologists of the center used the artificial zircon, apatite and magnetite distribution of the Coffdor deposit as an example to construct the three-dimensional structure of the deposits in the Murmansk region using artificial neural network. The model, the data used for the analysis, uses the chemical composition and mineral composition of the rock being drilled. This method was developed to optimize the geologist's work, using properly tuned “machine learning†(neural networks or other similar methods) and geostatistical deposits to create a 3D map of the deposit, which optimizes the mining and reduces costs Very necessary.
This new method has broad prospects for studying mineral resources in the Russian Arctic. The subject was supported by the Russian Science Foundation.