MLCOURSE.AI - Open Machine Learning Course By OpenDataScience.

Sept. 12, 2018, 10:31 p.m. By: Prakarsh Saxena

OpenDataScience that is one of the biggest IT communities in the world unites more than 15k Russian-speaking Data Scientists. Among dozens of ODS projects is organizing for researchers and practitioners the biggest Russian-speaking DS conference: DataFest. OpenDataScience also has its own educational mission that it has presented with the Open Machine Learning Course: MLCOURSE.AI.


MLCOURSE.AI is a 10-week open Machine Learning course by OpenDataScience launching on October 1, 2018. The course has been designed to perfectly balance theory and practice; therefore, each topic is followed by an assignment with a deadline in a week. You can also take part in several Kaggle Inclass competitions held during the course and do your own projects as well.

Throughout the course, they maintain a student rating that takes into account the credits scored by the students in assignments and Kaggle competitions. In accordance to the final rating, Top students will be listed on a special Wiki page.

So, what does the course comprise of?

The 10-week course comprises of the following:

  • Articles On Medium,

  • Assignments,

  • Tutorials,

  • Individual Projects,

  • Kaggle Inclass Competitions.

What are the Prerequisites for the course- Python, Math, DevOps?

  • Among prerequisites are knowledge of basic concepts from calculus, linear algebra, probability theory and statistics and Python programming skills.

  • Moving towards Python, interactive tutorials like CodeAcademy, DataCamp or DataQuest even will suffice.

  • Sufficient Skills in Docker, bash and GitHub are highly recommended as well.

Software Requirements:

Generally, the installation of the latest Anaconda 3 distribution will suffice as it contains NumPy, Pandas, Sklearn and lots of other libraries as well. However, some other packages like Xgboostand Vowpal Wabbit are also made use of.

Moving onto the Demo Assignments:

Every week in a new run of the course announcement of full assignments is made (October 1, 2018). While that takes place, an applicant can practice with the demo versions. Solutions to both demos as well as full versions are to be discussed in the upcoming run of the course:

  • Exploratory data analysis with Pandas, nbviewer, Kaggle Kernel

  • Analyzing cardiovascular disease data, nbviewer, Kaggle Kernel

  • Decision trees with a toy task and the UCI Adult dataset, nbviewer, Kaggle Kernel

  • Linear Regression as a problem in optimization, nbviewer, Kaggle Kernel

  • Logistic Regression and Random Forest in the credit scoring problem, nbviewer, Kaggle Kernel

  • Exploring OLS, Lasso and Random Forest in a regression task, nbviewer, Kaggle Kernel

  • Unupervised learning, nbviewer, Kaggle Kernel

  • Implementing online regressor, nbviewer, Kaggle Kernel

  • Time series analysis, nbviewer, Kaggle Kernel

  • Gradient boosting and flight delays. nbviewer, Kaggle Kernel

Resources Constituting The Course:

Here is the list of all resources that constitute the course:

  • OpenDataScience: All discussions are held in the #mlcourse_ai channel. The form to the invitation is mentioned in the end.

  • Kaggle Dataset: The whole course can be passed just in a browser.

  • Medium articles: Main content in English, not necessary if the user works with notebooks.

  • GitHub repository: Preferable if user is used to work with git & GitHub.

  • articles: Main content in Russian.

  • Youtube channel: Lectures and interviews with cool DS-guys. Yet Russian-only.

  • VK group: One more information mirror, Russian-only.

Kaggle competitions:

  • Catch Me If You Can: Intruder Detection with the help of Webpage Session Tracking. Kaggle Inclass.

  • How good is your Medium article? Kaggle Inclass


The course is completely free and the next session launches on October 1, 2018. So, Fill in the below form to participate and In September, you'll get an invitation to OpenDataScience Slack team!

Official Link:

Link To The Form: Click Here

For More Information: GitHub