The growing popularity of solutions based on artificial intelligence and machine learning has become a reason for increasing expertise in this area among developers and enthusiasts. Today you can find a lot of distance and full-time courses that allow you to understand the subtleties of developing projects with AI and ML, as well as gain practical skills in this area. We have compiled a list of the best educational programs in Russian and English that will give maximum benefit when mastering new technologies.
In the final project, you will be able to apply the gained knowledge to real data in ecommerce, social media, information search, and business intelligence. You will go through all stages of data analysis – starting from preparation to building a final model and assessing its quality.
Site: Yandex School of Data Analysis
Students attend lectures and seminars, work together on educational projects, the results of which often lead to scientific publications. Some students of the school can get an internship at Yandex, where they can apply their knowledge.
Fundamentals of artificial intelligence and machine learning for programmers.
Site: Artezio (AI lab in Nizhny Novgorod)
Price: free for company employees and interns
Duration: a series of lectures twice a year. Weekly classes and final work.
Courses for Artezio employees are conducted by Professor, PhD, Vladimir Krylov.
The Full Member of the Russian Academy of Engineering (RAE) Vladimir Krylov holds free classes on the essentials of AI and ML among programmers. Artezio has invested not only in creating a training course, but also in equipping and developing its own laboratory of artificial intelligence. Course participants receive not only theoretical, but also practical knowledge of working with AI and ML and implement their own educational projects in this area. The overwhelming majority of machine learning courses are aimed at gaining knowledge of how learning algorithms are built, the mathematical foundations of such algorithms, the use of machine learning for analytics. And only a narrow circle, in the majority closed corporate courses disclose the use of machine learning methods for software developers: architects, programmers.
Author: High School of Economics, Yandex School of Data Analysis
Duration: 7 weeks, 3-5 hours a week
This course tells mainly about the main types of machine learning tasks: classification, regression, and clustering. The teachers from Yandex and High School of Economics explain the basic methods and their features, teach to evaluate the quality of models and understand which problem each of them is suitable for. The program is designed for 7 weeks, but if you study hard, you can finish the course before September 1. The course is aimed at students who are familiar with Python, as its libraries such as Numpy, Pandas and Scikit-Learn are used.
Site: Great Learning Author: Great Learning Duration: 1,5 hours
A short course is aimed at those who are interested in machine learning, but do not yet know where to start. The program consists of 12 video lessons and explains what machine learning is and how an algorithm can be trained, introduces the basic terminology and methods, and provides practical exercises.
Site: Udemy Author: Kirill Eremenko, Hadelin de Ponteves, SuperDataScience Team
Duration: 41 hours of video lectures
The course is developed by two Data Scientists who explain a complicated theory, algorithms and programming using ML libraries. The course consists of 10 parts that deal with data processing, regression, classification, clustering, reinforcement learning, natural language processing, and deep learning. The course contains practical exercises and code templates for Python and R. Much attention is paid at choosing the right model for each type of task.
Site: Udemy Author: Jose Portilla
Duration: 21,5 hours of video lectures
The course helps you understand how to use Python to analyze data, create visualizations, and use machine learning algorithms. Such tools as NumPy, Seaborn, Matplotlib, Pandas, Scikit-Learn, Machine Learning, Plotly, Tensorflow are applied at the course. Also, students will be told about the natural language processing, artificial intelligence and deep learning.
The course involves the study of machine learning problems, the motivation to solve them, and examines the practical applications of these tasks. It consists of 9 training weeks. To successfully solve most of the problems contained in the tests, it is sufficient to master the material told in the lectures. The seminars also deal with more complex tasks that will be interesting for listeners who are already familiar with the fundamentals.
Author: Frank Kane
Duration: 12 hours of video lectures
The course covers the use of artificial intelligence and machine learning to solve business problems. Frank Kane worked for 9 years at Amazon and IMDb creating recommender systems. Each concept is described in simple language without complex mathematical terms. After the introduction, the use of Python code is demonstrated. The focus is on the practical understanding and application of machine learning algorithms. At the end of the course, students are offered to work on a final project to apply new knowledge.
Duration: 15 hours of video lectures
Google offers a quick and practical introduction to ML using API TensorFlow. The course includes a series of lessons with video lectures, real tasks and practical exercises. In total, students need to watch 25 lessons and complete 40 exercises. Interactive visualization is offered for all algorithms.
Duration: 2 weeks
Price: subscription to Coursera 3 039 ₽ a month
The teachers from Stanford University will tell you how to build the work of a team using ML. During two weeks, students will learn to find errors in the machine learning system, set priorities in the direction of work, and understand complex parts of machine learning, for example, invalid training data sets.
Author: Google Magenta
Duration: 5 sessions, 12 hours of work per session
Language: English, Russian subtitles
The course was created with the support of Google’s Magenta project, in which the company is trying to build a “creative computer”. The teachers talk about the main components of deep learning that are necessary for developing algorithms: convolutional networks, variational autocoders, generative adversarial networks, and recurrent neural networks. A major focus is also on creative neural networks. For example, working with an image and creating content that will match the aesthetics or content of another image.
Author: Carnegie Mellon University
Duration: 24 lectures for 1,5 hours each
Language: English, Russian subtitles
On YouTube, there are videos of a series of lectures by Larry Wasserman, Professor at the Department of Statistics and Machine Learning at Carnegie Mellon University. The course is aimed at people with advanced knowledge of mathematics and programming, as it focuses on the integration of statistics and machine learning. The pre-courses for the lectures are “Intermediate Statistical Theory” and “Introduction to Machine Learning”.
Duration: 6 weeks, 2 – 4 hours a week
Price: free, certificate for $99
The course is included in the Microsoft certification in the field of data science. It tells you how to create and work with machine learning models using Python, R and Azure Machine Learning. The teachers describe classification, regression in machine learning, controlled models, non-linear modeling systems, clustering, and recommender systems.
For those who prefer offline meetings, ITMO University will hold the Summer Machine Learning School from August 2-15 in St. Petersburg based on the Center for Speech Technologies. Students will gain practical experience in applying deep learning methods and algorithms to analyze audiovisual data for emotion recognition. Requirements to participants:
— Senior students
— Python knowledge
— Have experience using modern machine learning methods
— A great desire to develop in the field of audio and video analytics