We regularly conduct courses for a variety of people: for beginners , for developers with experience and heads of IT departments . Together with Dmitry Soshnikov ( shwars ) we launch AI School : a school where developers master all the main areas of artificial intelligence for a month, from chat bots and image recognition to working with texts and knowledge bases. The graduate of the AI ​​School will be able to immediately begin to apply AI technologies in his projects, and he will develop the first such project already within the school.

Dmitry Soshnikov is a school curator and evangelist of Microsoft, he holds a PhD in AI and teaches at HSE, MIPT and MAI. Other specialists in the field of artificial intelligence will also be speaking with him. In this article we will tell for whom this course was created, how it will be held, and what students will receive at the end of the course.

For whom AI School?

First of all, the school will be useful for developers - they will be able to apply their skills in their projects. Technical directors and IT project managers will understand the limitations of different approaches to AI and know the appropriate terminology in order to speak the same language with the development teams.

Business representatives will learn how to automate business processes. For an initial introduction to technology, short courses are suitable, such as AI Weekend . A long course allows you to understand all the possibilities and understand the technical details.

What will students get?

Students will learn to solve 5 main types of tasks:

• tasks solved by pre-trained cognitive services: recognition of faces, emotions, voices;
• creation of the simplest conversational AI;
• classical machine learning tasks: demand prediction, predictive analytics, etc .;
• work with images and video: classification and detection of objects;
• work with text and natural language: classification, generation, etc.

Most importantly, developers will not only begin to deeply understand the terminology of AI — reinforced learning, deep learning, etc. - but they will also learn how to apply the appropriate technologies to solve their problems. AI School will consider the main neural network architectures: convolutional networks for working with images and spatial data, recurrent networks for word and sequence processing, Siamese networks, autoencoders and generative contention networks - and try to use them as examples. Students of the course will gain the necessary base to further deepen knowledge in the field of AI on their own, reading specialized scientific articles.

Teachers will not only explain how to create software systems that are difficult to algorithmize, but also offer short practical tasks that will then go to homework.

Course speakers are professionals specializing in a particular field of artificial intelligence, technical directors of companies or AI start-ups: OpenWay, Nanosemantics and LoyaltyLab. The school will also include practicing Data Scientists, researchers in the field of AI, and teachers from leading technical universities - MIPT and Moscow State University.

Those who are already working with artificial intelligence or are thinking about how to apply it in their business will be able to get detailed advice from teachers on the introduction of AI in specific situations - for this there are breaks and question-answer sessions, as well as personal informal acquaintance.

Professions related to AI and Data Science are well paid and highly demanded. Knowledge in this area increases the value of a specialist.

What is the format of the course?

All courses in the Binary District are conducted offline. What it gives:

• the opportunity to talk with other participants of the course, get to know the teachers better and feel like a part of the community;
• the opportunity to consolidate skills: the course contains a lot of practical work, performing which in a team, students learn faster and get experience;
• The lively format disciplines and creates greater motivation to study the subject, which is difficult to maintain in the case of online courses.

Thanks to the lively format, students will get to know each other, will be able to share experiences and continue their communication after training.

The course program consists of 12 classes. Between classes there are always breaks - a day or two, during which you can independently repeat the theory and practice. The course will cover various approaches to AI; they will be considered as applied to specific tasks that arise in practice.

What you need to know before learning?

If we talk about the theoretical part, then there will be no proofs of theorems, but it is useful to remember what matrices are and how they are multiplied. The most difficult thing to face is the gradient function of several variables and partial derivatives. If you forget what it is, the teachers will remind you.

As for practice, most of the code examples in the course are provided in Python — the de facto standard in machine learning and AI. Therefore, for more productive learning, it is useful to at least read the Python code and be able to modify it slightly. Most of the tasks in the course of the course are solved by modifying the examples offered to students, so proficiency in the language is not required.

You can learn more about the program of the course and its teachers on the AI School page .