The science of neural networks. Live

There are 213 days left until the end of the year, so it's high time to start learning something new, for example, to plunge into the science of neural networks. Today, in one day, we will get acquainted with the installation of neural networks on the air, starting with simple architectures and ending with deep learning - networks with dozens and hundreds of layers. Also consider the convolutional networks used for image recognition, and recurrent networks for sequence analysis. And you will be able to train a neural network with us to solve non-trivial tasks - from recognizing handwritten numbers to recognizing cats in photos.

Live broadcast

This is a live broadcast of the intensive “Practical introduction to neural networks and deep learning”, which is held as part of “DevCon School: Technologies of the Future” . It will be interesting for developers who have not previously worked with neural networks, or have worked very little, and want to understand how to use neural networks and neural network frameworks to solve problems such as image recognition and sequence analysis.


')
For the training of neural networks we will use: Azure Notebooks, Microsoft Cognitive Toolkit (CNTK), virtual machines with GPU accelerators.

Software Requirements

To participate you will need your own laptop (running Windows or Mac OS X), on which are installed:

• Modern browser (Microsoft Edge, Internet Explorer, Safari).
• Windows Remote Desktop Connection (Remote Desktop) capable of connecting to virtual machines in Microsoft Azure (for Mac OS ).
• The bulk of the intensity will be held in a browser using Azure Notebooks or in a remote virtual machine in the Microsoft Azure cloud.
• To avoid problems, try to open any of the Azure Notebooks in advance from here .
• You also need a Microsoft Account - if you don’t have one, create a new mailbox for yourself on Outlook.com .
• We also recommend that you create a new mailbox for experiments in advance if you do not have a Microsoft Azure subscription, and you plan to use the trial subscription, which we will provide at the event.

Speakers

• Dmitry Soshnikov , Strategic Technologies Expert, Microsoft
• Mikhail Burtsev , Head of Laboratory of Neural Systems, MIPT
• Mikhail Kozlov , Head of Development, PIK Group
• Andrei Ustyuzhanin , Head of the Laboratory of the Faculty of Computer Science, HSE

Program

10:00 - 11:00 Opening ceremony, plenary report
11:00 - 11:30 Break
11:30 - 13:00 Practical introduction to neural networks and deep learning. Part 1
13:30 - 14:30 Break
14:30 - 16:30 Practical introduction to neural networks and deep learning. Part 2
16:30 - 17:00 Break
17:00 - 19:00 Practical introduction to neural networks and deep learning. Part 3
19:00 - 19:30 Closing ceremony

useful links

You can leave comments and ask questions to Microsoft experts in our YouTube channel , as well as in the chat for school participants in Telegram .

Also, we will conduct live broadcasts in the Microsoft Developer: VK and Facebook communities.