So you have come across the buzz about AI and saw a couple of videos from Simplilearn about “What is Machine Learning?”, or maybe, have read a few articles about Machine Learning, or you just came across it today. Anyhow, I would like to distill the road I took to start learning this field to you.

By far, the most common question that I am always asked by many friends, students, curiosity munchers is “How to get started with Machine Learning?”. Often, my answer would be the following: Do you want to be that person who has no idea about the math that goes behind an algorithm, but imports a library/framework and builds a model, or do you want to be the one who knows exactly what’s happening behind each epoch of an algorithm’s training and understand the constructs of the hypothesis? You have to choose one. Because, my friend, the first kind of people have failed after a few months eventually giving up(not all, but many). The latter are the ones who go about doing PhD’s or working on improving the existing algorithms or contribute to the community in some way possible.

Disclaimer

These are solely my opinions and the road that I have taken to learn Machine Learning, and I am still walking the way — There are a lot more steps I got to take too. I am in no way considering myself an expert in this vast field, nor do I claim that I am an elitist in teaching you the right steps to take. These are the ones I took, and possibly, would help you to walkthrough too. As the wise man said, “In an uncharted terrain, it’s best to take directions from someone who’s walked the way”.

The Why?

Before we get into the specific details on getting started with ML, the most challenging and intriguing question to ask yourself is “Why do you want to learn ML?”. Take some time off trying to answer this question — because it really determines whether this field is for you or not. If “Money” is all that you care about, there’s a tone of other high-paying jobs/roles that you can take up. But if you’ve got the passion to learn the Mathematics that goes behind an algorithm that can classify potential brain tumor from an image, contributing to society with your tech — then, you are probably at the right place. I don’t mean the “money” suckers aren’t supposed to work with ML, I am just apprehensive about the contributions that they’d make to the community or god forbid, nothing at all.

Please, Remove the Intimidation

When someone walks up to you and tells you that Machine Learning is all about Mathematics, that person didn’t lie. Trust me, Mathematics is all that there is(Not just in ML, but literally about anything that is scientific, computational in nature). The first step you need to take is: Removing the intimidation you have towards Mathematics. Many students really do hate Math for obvious reasons. In fact, I would blame the teachers — not the students. Math is such a wonderful happening in this cosmos, but teachers(not all) fail to reflect its beauty in the heart of the student. Thereby, the learner would never really appreciate its true beauty at all — Until someone who’s crazy about it like me comes up to teach you all of it in its most glorious form. Trust me on this, if you could obliterate the barrier that you have with Math, you can shine in any field. Curiosity and passion are all that it takes for that!

How?

What are the prerequisites? — You may ask. Addition, Subtraction, Multiplication and, Division along with Basic Linear Algebra, Basic Calculus, and familiarity with Vectors. If you have no idea about the latter, there’s a course I cannot stop recommending for everyone. This is the same course I took, and many Machine Learning aspirants begin there. It’s taught by Andrew Ng — Professor at Stanford University. https://www.coursera.org/learn/machine-learning

The above course is absolutely fantastic for beginners. Even for those who don’t have much familiarity with the topics I mentioned above — Because he teaches you the Math as well supplemented with additional resources. The course teaches you almost everything that is considered industry-standard knowledge for Machine Learning practitioners and the course is also available for free for everyone.

Mathematics

• Multivariate Calculus
• Statistics and Optimization Techniques
• Linear Algebra

To learn more along the way, you can get started and enroll in this specialization as well to learn the above-mentioned topics: https://www.coursera.org/specializations/mathematics-machine-learning

If you go through them, you would have a solid foundation in theoretical aspects of Machine Learning. I would be updating you with more courses that you can take up here.

I would also go about writing another one for the practical aspects of implementing ML. Such as learning to code(not specifically for ML, but as a skill), learning to read the documentation, and understanding a few fancy libraries/frameworks. Until then, keep learning. And always remember: “You can learn anything!”

For Precious, with Patience.