No abstract

No abstract

The article considers the possibilities of using the deep learning convolutional neural network ResNet in computer vision and image classification problems. The interpretation of the ResNet network and the datasets used for its training are presented, as well as a method for training a deep convolutional neural network with stochastic depth, which allows significantly reducing errors in the test sample.

The results of measurements of complex relative permittivity of bentonite and clayey sandstone with different degrees of salinity with low moisture are given in the range of temperatures –20° to  +105 °C at frequencies from 25 Hz to 1 GHz. It is shown, that even a small amount of water in sandy and sandy-argillaceous rocks causes an increase of the real part of complex relative permittivity at frequencies below 100 Hz. The explanation by linearly-broken dependence of refractive index on moisture is given at its small values. By a dielectric method it is shown that in the process of water film formation on the surface of a mineral, the water molecules binding energy changes. Big distinctions in low-frequency dielectric relaxation times testify to the change of binding energy of molecules of water on the surface of a mineral. Also dependences of relaxation times on temperature are various. The results of dielectric measurements showed a strong influence of the salt on the dielectric permittivity of the clay and clayey sandstone even at a low moisture level.

No abstract