How to learn Machine Learning in 2022

A step by step guide to getting into machine learning

Devansh
Geek Culture
7 min readJan 20, 2022

--

To help me understand you fill out this survey (anonymous)

As my content becomes more popular, I have more people reach out to me with a variety of questions. A common theme among them is from people looking to break into Machine Learning. More specifically, this is from people trying to teach themselves Machine Learning. They tell me about various struggles as they float around from courses, projects, and ebooks/guides without fully understanding what they’re doing and developing confidence in their ML. What’s more, they find themselves forgetting the basics, which further saps their confidence, and makes it hard to know how to proceed.

Since there are so many ways to do ML, people are trapped by the Paradox of Choice

You would think that with all the fantastic resources, it would be much easier to get started. There are tons of great libraries, tutorials, and courses for you to take. You can even get all kinds of certifications to know where you stand. However, the opposite is true. With all this information out there, it’s easy to get overwhelmed. I received the above message on Tuesday. Notice that despite being a Computer Science Major (the sender of the message is a Master’s student)at a well-known school, he still feels overwhelmed. This is due to the Paradox of Choice, a phenomenon where people feel overwhelmed by too many choices. Below is a well-known study.

All the options can really overwhelm people, causing them to get stuck and not even take the first step

In this article, I will give you a detailed step-by-step plan that will help you learn Machine Learning. There will be no paid courses/links so that this is accessible to more people. Make sure to take your time and really understand the basics before moving on. If you have any great resources, link them in the comments below/message me about them. Let’s help each other make it.

Step 1: Get familiar with Coding

This is a must. The absolute step 1. This should be a no-brainer, but people keep missing the extent to which coding is required. Yes, no code is a thing. Yes, packages like Tensorflow and Keras make creating models very easy. BUT they are a minuscule portion of the work you do. You’ll spend a lot of time going over the pipelines written by other people. You’ll have to look at the models published online with their documentation. You will spend a lot of time on very specific implementation details. All of this requires you to comfortable with coding. One of the best resources for this is freecodecamp’s Scientific Computing with Python. It covers most of the details and structures very well. Alongside it, get yourself familiar with recursion and backtracking. These are skills that are crucial in ML.

The course description. I love that they cover both DS and databases.

As you start getting familiar with creating your own classes, most simple automation, logging to files, and working with databases, you can proceed to step 2.

Step 2: Do the Math

A personal pet peeve of mine is when people confidently tell me that they don’t need math to do Deep Learning (Tensorflow is literally named after a math concept). I don’t know who started this silly rumor, but nothing could be further from the truth. Math is more than calculations. It teaches you how to think. It’s a language that you use to articulate exactly how you can frame and solve a problem. It is non-negotiable.

The reason I was able to come up with novel and effective solutions was because of my math training.

Now that we are on the same page, let’s talk about how much you need. It’s not a lot. As long as you get familiar with mathematical thinking, and understand the core concepts, you will be effective. Naturally, as you try to get better, Math will be more and more important. Make sure you watch the following video till the end. It goes into detail on the four math topics you need (and how much you need). It is a guide to help you get to a good baseline level. If you decide to proceed beyond that, it will only help.

As for the sources you can learn Math, Khan Academy is king. Fantastic courses, tons of problems, and a great system. Professor Leonard has good upper-level calculus. YouTube is a great place where you can learn and improve your mathematical skills.

Step 3: Basic ML

Once you start to get a mathematical intuition, start looking into understanding basic Machine Learning. ThreeBlueOneBrown has a great playlist on Neural Networks. StatsQuests and RitvikMath are two fantastic channels that talk about a lot of Machine Learning/Data Science related concepts in a clear manner. There’s also a certain YouTuber that explains different Machine Learning ideas and concepts in a clear and applicable manner (wink wink).

This video explains how to design and build ML projects to maximize your learning and employability.

If you’ve kept up with the basics, this is when you will notice how useful the basics are for ML. You will be able to read the documentation, go through people’s GitHub projects, and mostly understand the resources. This is when you should get into developing projects. The above video is a guide on exactly how you should design your ML projects to have the greatest carryover to practical learning.

At this point, you’re still going to be a relative beginner. However, as you start to engage with ML channels and learn more about things, you will start to notice patterns and thoughts. And now you’re competent at Machine Learning. Now is when your growth will start to accelerate. How? Look at the next step.

Step 4: Deep dive deeper into the Papers/ML community

Up to this point, we have focused on equipping you with the right tools. You have developed the ability to understand Machine Learning both theoretically (through your mathematical grounding), and implementation-wise (through coding). You should be able to explain the idea and concept behind the common algorithms and implementations.

Variants of this quote are around the internet. This is what we have focused on.

The next step comes in with engaging with more complex literature. This can be disheartening at first. You will open papers, and see papers and notation that you are confused about. The ML talks shared will seem like 5 minutes of English and 40 minutes of Jargon. That’s okay. It’s a gradual process. The more papers you read/learn about, the more you will be able to understand. In the beginning, you can focus on understanding just one or two things from each paper (my annotated papers can help with that). The more you do, the better you will get at gaining insight from them. This article goes into detail about how you should interact with these complex ML papers to boost your Machine Learning skills. It is one of my best-received articles to date.

Once you get to this stage, it’s a lifelong process. You’ll keep learning from the tons of smart people that share their knowledge online. Some good starting places are Henry AI Labs, Yannic Kilcher, Robert Miles, Mathematical Monk, and Primer. You should also attempt harder projects such as the one detailed here.

For best results, make sure you subscribe to my daily newsletter, Coding Interviews Made Simple. It covers topics in Algorithm Design, Math, Recent Events in Tech, Software Engineering, and much more to make you a better developer. I am currently running a 20% discount for a WHOLE YEAR, so make sure to check it out.

I created Coding Interviews Made Simple using new techniques discovered through tutoring multiple people into top tech firms. The newsletter is designed to help you succeed, saving you from hours wasted on the Leetcode grind. You can read the FAQs and find out more here

Feel free to reach out if you have any interesting jobs/projects/ideas for me as well. Always happy to hear you out.For monetary support of my work following are my Venmo and Paypal. Any amount is appreciated and helps a lot. Donations unlock exclusive content such as paper analysis, special code, consultations, and specific coaching:

Venmo: https://account.venmo.com/u/FNU-Devansh

Paypal: paypal.me/ISeeThings

Reach out to me

You can reach out to me on any of the platforms, or check out any of my other content. If you’d like to discuss tutoring, text me on LinkedIn, IG, or Twitter. Check out the free Robinhood referral link. We both get a free stock (you don’t have to put any money), and there is no risk to you. So not using it is just losing free money.

Free Weekly Summary of the important updates in Machine Learning(sponsored)- https://lnkd.in/gCFTuivn

Check out my other articles on Medium. : https://rb.gy/zn1aiu

My YouTube: https://rb.gy/88iwdd

Reach out to me on LinkedIn. Let’s connect: https://rb.gy/m5ok2y

My Instagram: https://rb.gy/gmvuy9

My Twitter: https://twitter.com/Machine01776819

If you’re preparing for coding/technical interviews: https://codinginterviewsmadesimple.substack.com/

Get a free stock on Robinhood: https://join.robinhood.com/fnud75

--

--

Devansh
Geek Culture

Writing about AI, Math, the Tech Industry and whatever else interests me. Join my cult to gain inner peace and to support my crippling chocolate milk addiction