A hands-on roadmap to learn Data-Science, Machine-Learning and MLOps. This roadmap is more practical than theoratical (great news!); So it's best to do more research as you reach sections of this roadmap.
- Push-ups with Python
- SQL is your friend
- Version Control with Git
- The Era of Data Science
- Building Blocks
- A Gentle Approach to Machine-Learning
- Get to Know the Concepts
- Q&A Session
- Dive Even Deeper
- Docker in Action
How do functions break-up? Well, they stop calling each other.
You can use almost any programming language you'd like, but Python is the widest used language among Data Scientists and Machine Learning Engineers. It's also super easy to master, has thousands of libraries and one og the biggest communities!
Read the complete roadmap here!
A SQL query goes into a bar, walks up to two tables and asks: "Mind if I join you?"
SQL is the backbone of databases and as a Data Scientist, you get to work with a lot of complex queries which can and will blow your mind. Better come prepared...
Read the complete roadmap here!
The problem with git jokes is that everyone has their own version.
You may not need to work with Git on your first day, but as your models get more complex and you want to use your models in production, Git is your guardian angle!
Read the complete roadmap here!
"The first rule of data science is: don’t ask how to define data science!"
Here comes the good stuff. Time to get your hands dirty Deep Learning Models!
Read the complete roadmap here!
What if we put Docker inside of Docker?
- Why even bother with Docker?
- get-started
- Components in detail:
Checkpoint #9: Dockerize your Reddit scraper bot and deploy it on Heroku (run it locally first, remember baby steps!).
Final Note: Getting comfortable with Docker can take some time and practice. However, Docker is the gateway to MLOps which plays an important role in deploying Machine-Learning models!
We are here to share knowledge, don't by shy (: