A lot of data science courses claim to be free, but most are not. Many, such as Coursera, let you sign up for free for a trial period of about a week. After that, you typically have to pay about $49 per month.
However, some are completely free unless you want a certificate. EDx.org, for example, offers dozens of data science-related courses. EDx.org offers almost every data science topic you can think of.
These include several different Introductions to data science, linear algebra for data science (which I touched on in one of my first blog posts, as well as various topic in machine learning, deep learning, data mining, big data analytics, and many others. Some of the courses are available through Microsoft, and other courses are offered that include the Data Science and Machine Learning Capstone Project through IBM, Data Science Visualization through Harvard University, Python for Data Science, through UC San Diego. Other Python courses for data science are from IBM and Microsoft.
Some of these are free and some only have a short trial period. This depends on who produces the courses. I have found most of them to be free for an extending period of several months.
Exceptional Data Science Courses
In 2021, I ran across FutureLearn which offers a series of exceptional data science courses along with a large series of course offerings about science, engineering, math. FutureLearn offers a free 7-day trial period.
Open Source Data Science Masters
KD Knuggets has published a listing of courses for data science that they call an open-source data science masters. It is not really a masters degree, but the courses are great. Most of them are free except for those from Coursera, which only have a 7-day trial period.
Introduction to Machine Learning for Coders
The University of San Francisco developed a course called Introduction to Machine Learning for Coders.
This course provides about 24 hours of lessons. It has a prerequisit of a minimum of at least a year of coding experience. The University of San Francisco course is based on lessons recorded for the Masters of Science in Data Science program.
FastAI, which was also created at the University of San Francisco, is free if you have the right hardware on your desktop computer. It requires an i7 Intel processor with a relatively recent model of an Nvidia GPU. This configuration is virtually impossible on Mac and can only use a Windows pc. I have watched several videos of this course and found it to be fascinating. In the future, I intend to get a PC with this configuration for just this purpose. If you do not have the right hardware, they recommend using cloud access, which is not free.