Using the Artificial Neural Networks with Seismic Faces Analysis for Reservoir Modeling

In this paper we shows the method of the seismic facieses analysis by poststack seismic<br>attributes. There are used neural networks (Kohonnen’s SOFM) for distinguish of seismic<br>facieses in area version of 2D seismic data. The methods are based on the analysis of digital<br>map of seismic attributes in their representation by matrices. The usage of a representation by<br>matrices of map the seismic data attributes 2D gives the wide possibilities of well-known<br>methods of mathematical statistics apply. From technological schedule the data representation<br>like a three-dimensional matrices allows to use the popular DBMS and electronic tables for<br>initial data preparation and analysis of outcomes.