Modern State of Artificial Intelligence

• Math basics of AI

The course provides the mathematical foundation necessary for further work in artificial intelligence. It includes the fundamentals of mathematical analysis, linear algebra, optimization methods, and discrete mathematics.

• Probability Theory and Statistics

The course is intended for masters of mathematics interested in modern discrete mathematics methods and applications of probability in computer science. The course includes all the basic definitions and statements of probability theory: Kolmogorov’s axiomatics, distributions, random variables and vectors, expectation, probability inequalities, laws of large numbers and the central limit theorem.

• Software Development with Python

This course focuses on the fundamentals of software engineering. Good design is an important part of any project. This course covers the basics of the Python programming language, basic concepts, and language constructs. It also provides tools for using the Python language in complex projects. Students will gain insight into proper code design, code base maintenance, and integration of your applications with others.

• Algorithms and Data Structures

This course is designed to provide a theoretical foundation in computer science. Unlike courses in mathematical programming, it focuses on concepts and ideas rather than detailed proofs. The main goal is to provide an intuitive understanding of what constitutes an algorithmic problem, which problems can be solved in principle, and which problems can be solved efficiently.

• Machine Learning

This course is designed to introduce students to the current state of the art in machine learning and artificial intelligence. It combines theoretical foundations of machine learning algorithms with extensive practical exercises. The course covers material from classical algorithms to deep learning approaches and recent achievements in the field of artificial intelligence.

• Applied Statistics and Data Analysis

The course contains basic information from mathematical statistics, which will be used and supplemented in the future. The article considers classical probabilistic models of decision making about classes of observed objects according to the values of their attributes (models of classification or choice of hypotheses); it is assumed that the distributions of a feature for each class of objects are known exactly or with accuracy up to type. Goodness-of-fit criteria are discussed as a tool for testing the reliability of hypotheses and the problem of estimating distributions (in particular, the problem of so-called parametric estimation).
Methods of nonparametric estimation of distributions, which require much less a priori information about their properties, are discussed separately. The course also includes basic information on regression analysis, which is used to identify and evaluate the probabilistic relationships between the random variables under study.

• Software Development & Data Storages

This course focuses on the fundamentals of software engineering. Proper design is an important part of any project. This course covers the basics of the Python programming language, basic concepts and language constructs. Along with this, this course provides tools for using Python programming language in complex projects. You will gain insight into the correct design of the code, maintaining the codebase and integrating your applications with others.

• Recommender Systems

In today’s world there is a lot of different content: videos, books, music, articles. Due to the abundance of information it is difficult for a person to choose what is interesting for one from the variety of options. Such problems are solved by recommender systems. They are present in a wide range of services: online shops, video, news feeds etc. Although we do not always notice it, recommendation algorithms have become an integral part of our lives. In this course, students will examine the classical algorithms of recommendations: from matrix factorization to modern neural network approaches.

• Computer Vision

Modern computer vision is almost entirely based on deep learning. However, an understanding of classical approaches is essential for a thorough understanding of the field. Both novel approaches and biological inspirations of the field are covered in this course. Practical exercises will help students to acquire the relevant skills and to improve their general understanding of the field of AI.

• Natural Language Processing

In this course, we will start from the basics and quickly dive into the latest results in Natural Language Processing, focusing on novel approaches and applied techniques. This course tends to develop both practical skills and theoretical background to provide students with thorough theoretical knowledge and the ability to work independently on NLP projects.

• Reinforcement Learning

This course follows the Machine Learning course and aims to introduce students to the current state of Reinforcement Learning. The biological inspiration and theoretical foundations of Reinforcement Learning algorithms are reinforced with extensive hands-on exercises. The course covers material from Genetic Algorithms to the latest achievements in Reinforcement Learning approaches and narrows the gap between different subfields of Artificial Intelligence.

• ML Ops

This course covers the basics of design and development of modern MLOps platforms. From initial project scoping to exploratory analysis and data engineering, from feature engineering to model training, to ML pipelines and CI/CD automation, productization, monitoring and operations. As the course main deliverable, you’re required to implement an end-to-end ML project with the main components of MLOps, such as reproducible ML pipelines, scalable data processing and model training, model registry, experiment tracking, observability, and event-driven workflows.