Neural Networs technologies and theirs application in oil prospecting seismology

There the results of using the neuronet models (Kohonen net and 3- layer perceptron) for the solving of<br>clusterization, classification tasks, also for the making the regression analysis of objects on Ю0 horizon, of<br>studing seismic block are presented. As the sourse information we use 3D seismic trasses of studing horizon<br>and debits significance of some wells. The solving of tasks is making with the accounting of mutiextrmalness<br>of mistake function.