The ideal way to learn about data (2022)

The definitive roadmap for 1st and early 2nd year students!

Hi folks! This blog is about how I would have learnt data science and related fields if I had to start over. Please follow this guide if you are in your 1st or 2nd year and want a deep understanding of the end-to-end design of data applications in any tech company so as to have a bright career in this field! I guarantee that if you follow along this path, you’ll possess the knowledge of someone with at least 3 years in this industry.

Disclaimer-

1. This is NOT a free guide. To the best of my estimates, following the entire roadmap will cost you close to 60,000 INR (excluding GST). However, trust me, this investment will be a much better one than a so-called certification training from a coaching institute which would cost you as much, if not more.
2. This is NOT a guaranteed path to crack a 30–40 LPA job straight out of college. I can only assure in-depth knowledge but not such a salary from this course. If you want that, feel free to tread the competitive programming and open-source path.
3. I am in no way affiliated with either of these courses. I am just a student and know neither any of the institutes nor their founders. 🙂

If you are still interested, please read on and follow the order in which it is the roadmap is laid.

Python-

I’d like to point out 3 resources. You can use any or all depending on your time constraints-

1. Python for Everybody- Charles Severance on Coursera- Frankly, the first course for 60–70% of Python developers out there. You’ll be taught by an amazingly cool professor from the University of Michigan.
2. Python playlist by Dhaval Patel- You’ll be learning from a software engineer who has worked with Nvidia and Bloomberg. Consider following his channel too if you want to get into this field.
3. Python playlist by Tim- Young software engineer with whom most of you can resonate. He has 3 playlists- beginner, intermediate and advanced. You can also consider building some cool Python projects from his channel. Worth subscribing!

P.S- If you have time, consider checking out Flask/Django frameworks from Tim or Prof. Charles Severance (more theory-intensive). But you can skip it for now.

SQL-

Learn this from freeCodeCamp or Mosh. The main key here is to practice. Do that from Leetcode, Hackerrank and SQLZoo.

Data Structures and Algorithms-

Might sound counter-intuitive but I suggest not doing this in Python. Choose C++ or Java and practice 1–2 Leetcode easy to mediums daily just in case you have to fall back to the traditional software engineering job. Trust me, you’ll get a gazillion of resources on YouTube for this. 😂

Data Engineering-

Yup, you heard that right! You will have a better foundation for Data Science and ML if you start with Data Engineering. TrendyTech is a good course to take in this field. Truth be told, there aren’t a lot of quality, practical courses in this field anyway.

Note- Data Engineering is a booming field. If you love the course, your journey ends here (well till college at least), and there’s no need to read further. Complete the course, do as many internships as you can. Skilled data engineers are paid as much, if not higher than data scientists and there are very few skilled professionals in this field yet the demand is ever-growing. Besides, this field is more welcoming to freshers with just a Bachelors’ degree than Machine Learning. So the path’s pretty rosy.

Data Science and Machine Learning-

Now, here’s the juicy part of the post. To spit facts, AppliedAI Course is hands down the best course out there in this field. The instructor, Srikanth Verma sir (with rich experience of having worked in Amazon and Yahoo in machine learning roles) is the best ML instructor in India I’ve seen and his teaching skills is on par (if not, at times better, XD) with the instructors at Stanford, MIT, etc. He will make you fall in love with machine learning with the perfect blend of theory and practical nature of his lectures.

Note-

1. There are 30 assignments in the course which are quite tough nuts to crack. Good luck with that.
2. You’ll get some live sessions along with the course lectures. You can ignore the ones for experienced folks like those of ML system design etc., for now. There’ll be some code-walkthroughs and interview questions like sessions which you should watch.
3. You’ll be having a good portfolio of projects by the end of this course. Consider deploying the projects using methods discussed by Srikanth Sir in the live videos (Flask, Streamlit, Sagemaker, KubeFlow, etc). This helped me a lot.
4. iNeuron too has a data science course. It is more focused on coding applications rather than theory. That’s another alternative for you.
5. This course covers 70 percent of data science in the real world. For the rest, read on.

Practical Data Science-

Try finishing many courses from LogikBot to gain experience on what kind of data science is actually done in the real world. Reading e-books by Jason Brownlee is another fantastic way of doing the same. Basically understanding the coding aspects.

Conclusion-

Whoa! That’s a lot of time involvement. I’d estimate at around 2 years for a relatively bright student. If you somehow still have time, consider learning Flask/Django and watch ML system design videos by AppliedAI particularly if you are aiming for Machine Learning Engineer roles.

If you have completed the roadmap, are on the journey, or planning to get started feel free to connect! Please ask queries in the comments section. And in the end, be generous in your claps, and do follow me for more!