Speaker: Karandashev Iakov, Leading research assitant at Centre of optical and neural technologies of SRISA RAS, Associate Professor of the S.M. Nikol'skii Mathematical Institute, RUDN University.
Topic: Neural networks: architectures and opportunities of deep learning
Annotation. The report will tell about neural networks, algorithms underlying the training of neural networks, modern frameworks that are used to implement neural networks, how much data is needed for training, how and in what areas they work. Great attention will be paid to various deep learning architectures, such as: classical multilayer perceptrons, convolutional neural networks, GANs, autoencoders, recurrent neural networks, etc. Finally, perspective opportunities of using neural networks in the near future will be briefly outlined.