A Roadmap to Becoming a Self-Taught Data Scientist

Let’s say you have set a 6-month goal learn to data science and you are a complete newbie, here is the road map I have created to take you there from the scratch. It covers python programming, data science intuition and solving real world problems:

  1. Dataquest (2 months) [Hands-on Tutorial & Coding]- For hands-on python programming from the scratch. Good thing is that you’d get to work on real life projects and write the codes yourself. You’d start from the scratch with python and then learn to clean and manipulate your data. All these are done with real world data and it gives you the intuition to approach any data problem. Highly recommended.
  2. Automate The Boring Stuff [Book] (2 weeks) – This book is a good material that got me started with practical applications of Python. It teaches you how to use python to automate manual tasks like renaming files, moving documents and creating a csv file for data manipulation.
  3. Machine Learning AZ™: Hands-On Python & R In Data Science – Udemy (2 weeks) – [Video] The reviews are great and it got me started on solving a real machine learning problem. Its the best video tutorial I have taken on machine learning. At the end of this tutorial you will understand clearly all most of machine learning codes you come across that once looked like Greek. And Its the most upvoted machine learning course on Udemy.
  4. Andrew Ng’s Machine Learning Course – Coursera (2 months) – [Video] – This course is highly recommended and will deepen your knowledge and fill in the gaps that were not covered in A-Z machine learning. This course focus on the underlying concepts, the maths and the crux of machine learning. You will learning how to calculate the loss function by hand and what cost function means. Very intuitive and stimulating. Its the most upvoted machine learning course on Coursera.
  5. Kaggle (1 month) – This is the most important part as you will begin to work on real problems. I strongly suggest you replicate already solved problems and kernels before you work on new ones. You’d have a good enough knowledge of data science by replicating all the projects in these materials. Kaggle is free.

According to Wikipedia, a technology roadmap is a flexible planning technique to support strategic and long-range planning, by matching short-term and long-term goals with specific technology solutions. So this would be updated when I deem it fit.

4 thoughts on “A Roadmap to Becoming a Self-Taught Data Scientist”

  1. Great list. Thank you for sharing it. I am learning python right now as a way to become more creative and learn more about programming and the internet. The Roadmap here is very useful for a newbie like me. In fact the Automate the boring stuff is better than a coursera course on Python I took recently. It just shows that newbies cannot really evaluate what is worth it. Wish there was a way to know what would give you the best outcome for the time you put in.

    Keep up the page. I am slowly getting into it. The fact that you are based in Kenya is great as well.

    • Hi Rajana,

      Thanks for the response and its been a while since I put up a blogpost. So sorry my response is coming late.

      How is your journey coming along? Do you still study with Automate The Boring Stuff? Do you have any project(s) in mind?

      Let me know and please do have a wonderful time.

      Hopefully will continue to write from now.


  2. Fantastic article ! You have made some very astute statements and I appreciate the the effort you have put into your writing. Its clear that you know what you are writing about. I am excited to read more of your sites content.

  3. dataquest.io started adding a bunch of material after this was written. It might take 300-400 hours to go through everything now.


Leave a comment