# How to Self-Study to Be a Data Scientist – Adapted Compilation

My name is Ira, I blog about mathematics, products and ML, do projects in DS / ML, as well as courses on creating ML projects. From my blog audience and friends I often get questions like “Is it worth spending money on this or that expensive course with the codename” Become a DS in two months “or still learn to be a data scientist on your own and for free, and in this case, *where to begin?”*

Recently, I have collected an answer to this question in detail, in my opinion, to share it with a wide audience. Not all paid courses are bad (although most – yes – due to selection mechanisms, sales and weak programs), but I will write about them separately. It seems to me that you need to take into account more personalized parameters in order to correctly choose a good course for the money.

First, the list includes

**free online courses and self-study resources**, which she passed herself or is advised by the professional community of analysts and data Scientists (often mentioned in ODS.ai), where you do not need to go through some kind of selection or be limited to offline visits.Second, of course, it is

**not a complete list of online courses**which you can find, but it contains the best courses from strong math and Computer Science schools in the world and other common resources among the pros from what I could filter to my taste.Thirdly, I will start with a short list, with which, it seems to me, it is worth starting learning the subject, and it goes immediately into the next paragraph.

### Where do I recommend you start in order to spend time efficiently?

Assuming the aspiring Data Scientist has already completed Harvard-style prep courses CS50 on the basics of programming, pythontutor.ru or course on Stepik from the Institute of Bioinformatics, I will advise a few steps that should become a solid base. Further links are all complete to make it easier to copy:

Register in the most popular professional slack community in the CIS Open Data Science ODS.ai, join as many chats as possible, including about mentoring, training and careers and communicate with locals in order to broaden your horizons regarding employers, interview requirements, positions and their differences in different companies, etc., find good mentors, maybe … there are such in the community!

Take an introductory course on Math and Python for Data Analysis on Coursera — paid, inexpensive, and good. www.coursera.org/learn/mathematics-and-python

Programming – go to leetcode.com all relevant exercises: these are free or inexpensive in the premium version at a price / quality ratio (it includes exercises for interviews in FAANG).

Pass the mlcourse.ai Is an open source machine learning course from ODS. The authors were able to develop a machine learning course with a balance between theory and practice, when in a lecture you analyze mathematics in sufficient detail, and then practice first in a notebook, then in Kaggle.

To learn how to solve various problems and optimize code – participate in competitions in data analysis and machine learning on the platform kaggle.com…

### Math for DS / ML

Nice Stanford Course on DS “Introduction to Statistics” www.coursera.org/learn/stanford-statistics

A short interactive course on probability theory and mathematical statistics “Seeing Theory” seeing-theory.brown.edu/

A good introductory course in mathematics for data analysis, more voluminous “Specialization Mathematics for data analysis: You can only listen to an interesting topic: discrete mathematics / linear algebra / mathematical analysis / probability theory. www.coursera.org/specializations/maths-for-data-analysis

### Python & SQL for DS / ML

The above free Python trainer from scratch: pythontutor.ru/

Excellent course on DS tools from IBM “Data Science Fundamentals with Python and SQL Specialization” www.coursera.org/specializations/data-science-fundamentals-python-sql

The above Russian course on Python and Mathematics (paid, inexpensive and good) “Mathematics and Python for Data Analysis (Coursera)” www.coursera.org/learn/mathematics-and-python

Mentioned above https://leetcode.com/: complete all relevant exercises, these are free or inexpensive in the premium version in terms of price / quality ratio of simulators (it also includes exercises for interviews in FAANG).

### ML beginner courses

No matter how scolded this course is because of the outdated Octave programming language (which is written in Matlab), for my taste this is the simplest and most understandable ML course so far. Machine Learning (Coursera) https://www.coursera.org/learn/machine-learning – Stanford Machine Learning Course by Andrew Ng

mlcourse.ai Is an open source machine learning course from ODS. The authors were able to develop a machine learning course with a balance between theory and practice, when in a lecture you analyze mathematics in sufficient detail, and then practice first in a notebook, then in Kaggle.

### More advanced ML courses

If you want to immerse yourself in mathematical proofs of machine learning methods, then there are excellent ShAD lectures by K.V. Vorontsov: playlist “Machine Learning Course 2019” on the YouTube channel “Computer Science”, www.youtube.com/watc? v = SZkrxWhI5qM & list = PLJOzdkh8T5krxc4HsHbB8g8f0hu7973fK & index = 2

Also good is the annual Harvard course Advanced Topics in Data Science CS109B. harvard-iacs.github.io/2020-CS109B/

Or a course on advanced algorithms Advanced ML from HSE: “Specialization Advanced Machine Learning” www.coursera.org/specializations/aml

### Deep learning

(It seems to me that you can take one course from the list, and look at the rest at the subject of additions)

Recommended Stanford DL Course “CS231n: Convolutional Neural Networks for Visual Recognition” cs231n.github.io/

Good course from Carnegie Mellon University “11-785 Introduction to Deep Learning” deeplearning.cs.cmu.edu/F21/index.html

MIT Course: “Practical Deep Learning for Coders” https://course.fast.ai/

SHAD deep learning course is available on github: “Practical_DL” github.com/yandexdataschool/Practical_DL

Free classroom courses from MIPT: dlschool.org/

There is also a DL course at ODS.ai, which is also advised to take in the community itself: “Deep Learning on the fingers” dlcourse.ai/

### Natural Language Processing

### Reinforcement Learning & Self-driving cars

### Data Engineering & MLOps

Free DE course from Dmitry Anoshin, data engineer from Microsoft, ex-Amazon: Getting start with Data Engineering and Analytics https://datalearn.ru/ (course is being prepared in progress)

### Competitions

Well, and to learn how to solve various problems and optimize code – participate in Kaggle. https://kaggle.com/

Besides Kaggle, there are several more competitions:

Success in self-preparation!