Hey there,

If you’ve clicked on this post then you’re probably deciding on learning Machine Learning. If you haven’t already decided, or are still feeling confused, I want to tell you in advance it’s going to be daunting.

Those of you who do end up taking the leap of faith will probably get more out of this journey than anyone else. It will not only open your mind to a whole new way of thinking, it will show you a world of possibilities and what you could do to apply this learning to your world for the better in more creative ways.

The problem of beginning with Machine Learning right now is there are decades of research and multiple entry points one could start from. It’s a bit like Thomas Edisons 1000 failed attempts to invent the light bulb (or finding 999 ways not to make one). AI has gone through decades of failed attempts (AI Winters), but only now has it really propelled forward. So do you start at the 1st attempt, the 79th attempt, or the 999th attempt? Lucky for you, we can get straight to the point where everything is working (or at least seems to be working… for now).

The first place to start irrespective of where you are in your learning curve is by watching Frank Chen’s AI Primer from Andreessen Horowitz. No questions asked, just watch it. Frank Chen does an amazing job in the primer simplifying Artificial Intelligence for everybody.

This is also a great point to self-evaluate if you truly feel passionate about getting into the weeds with AI. If AI is only something you’d like to read up so you can appear smart during meetings I suggest this budding reading list from Harvard Business Review. But if you want to dive deep into AI and pick up the tricks of the trade then keep reading.

The first step to learning something new is assessing what you already know and what you can readily translate. To make this simple I’ve categorized three possible stages of where you may likely be at. These stages are dependent on two of the most fundamental skills that will help you acquire the knowledge you need to getting started with Machine Learning (wait for it) — Math and Programming. Now, now, don’t give up already. A relatively good understanding of Math and broad understanding of how to write trivial programs can get you real far.

Absolute Beginners (approximately 20 weeks)

I’m going to confess — I had to think real hard before making this recommendation, and I’m making it knowing fully how valuable the next 20 weeks are going to be for an absolute beginner. A strong foundation is necessary to ensure longevity in your skills and success in pursuing Machine Learning (and to go beyond). Knowing you aren’t comfortable with Math (and there could be plenty reasons for that — bad teachers et all) or Programming means you’ll have to roll up your sleeves and do the due diligence. Challenge yourself to put in the hard work and you will come out strong.

A big part of this recommendation will require you to sign up on Khan Academy (don’t worry it’s completely free and a whole lot more fun). I have a graduate degree in Computer Science Engineering which required me to do complex math. I can wholeheartedly say Khan Academy has given me more confidence to do Math than all of my teachers combined in all my academic career.

Algorithms — 1 week, free

We start with algorithms, and this should go pretty quickly. A good understanding of algorithms and a little proficiency can help you quickly pick up the skills and even master Deep Learning Algorithms down the line. The recent success in the field of AI is credited by one part to better training algorithms. Which is why starting here becomes almost crucial.

Python Basics on Mimo — 5 hours, free for the first 3 days, $39 after a 20% discount coupon which they will send you after the trial period

I love this recommendation. Mimo is a great app if you want to learn how to program in Python. After a long dormant period of no-coding I’ve rekindled my love for writing code through Mimo. Use it on the subway or in an Uber, you can get quickly from barely knowing how to code to writing cool apps. The app suggests 4.5 hours until you finish core concepts and that may really be all you need to get started with Machine Learning.

Algebra — 10 weeks, free

This is NOT something you can ignore. A good understanding of Linear Algebra is necessary for Machine Learning. You can self-pace this section based on how confident you get. Without a doubt, I would recommend completing the entire course and winning all those cool badges on Khan Academy.

Statistics and Probability — 10 weeks, free

Finally, you will need a solid footing in Statistics and Probability. After all, building general intelligence into a machine is all about being able to predict the probability or likeliness of things.

IF

You have successfully completed this stage of Algorithms, Python Basics, Algebra, and Statistics and Probability give yourself a huge round of APPLAUSE. That could not have been easy, but rest assured it’s going to be completely worth it. You are now a ‘Positive Beginner’. Not bad eh, you’ve gone from an ‘Absolute Beginner’ to a ‘Positive Beginner’ in just 20 weeks. You have nothing but your passion and dedication to thank. Proceed to the ‘Positive Beginner’ section.

IF

You would like to really get your fundamentals right so you can arrive at the Colosseum of AI like a gladiator, do the Pro section below.

(Go Pro)

Python (The Hard Way) — Self paced, free

Calculus — 15 weeks

Differential Equations — 3 weeks

Relative Beginners (approximately 3 Months)

The best place for a ‘Relative Beginner’ to start is by training with a pioneer — Andrew Ng. Andrew’s course on Coursera takes approximately 11 weeks to complete and is highly recommended by the folks who’ve taken it. You can even get a Coursera certificate by the end of the course for $79 if you’d like to rack up those AI credits.

Andrew Ng’s Machine Learning Course on Coursera — free, $79 for a certificate from Coursera credited by Stanford Online

Prerequisites for this course:

  • Knowledge of basic computer science principles and skills, at a level sufficient to write a reasonably non-trivial computer program.
  • Familiarity with the basic probability theory.
  • Familiarity with the basic linear algebra.

IF

At any given point you feel like what Andrew is teaching is too complex, this may be a good time to rethink where you are on the learning curve. Andrew is conducting this course with an assumption you fulfill all the prerequisites. I too have had to occasionally look back to Khan Academy at times to recollect forgotten concepts in Math. If this is something you’re comfortable doing, going back and forth, and are willing to take that effort then go for it.

IF

You have completed this section proceed to the ‘Positive Beginner’ section below. Note: The section below should take you a lot less time to complete given you already have had a great start.

Positive Beginners (approximately 4 Months)

Congratulations! You’re ready to become a Jedi Master in Artificial Intelligence. The ‘Introduction to Artificial Intelligence’ course on Udacity is conducted by Sebastian Thrun (ex-CEO and Cofounder of Udacity, ex-Google Fellow, Stanford Computer Science professor) and Peter Norwig (Director of Research at Google). You’ll be training with the best, through a unique interactive video experience on Udacity. This is a course put together by Google (absolutely free) and while it may seem like it was conducted in the 90s everything they’re teaching is highly relevant and gives you a good understanding of Artificial Intelligence in general. They even touch up on some essential topics as you progress like Game Theory, Computer Vision, Robotics, and Natural Language Processing. You’ll come out learning a lot more than you thought you would.

Introduction to Artificial Intelligence — free

Where to go from here?

Take a moment to congratulate yourself on everything you’ve accomplished until now. If you’ve transitioned between the three stages and successfully completed the course on Udacity you deserve a break. That must have been a lot of hard work but I’m sure it was fun and exciting throughout.

Recommended light reading:

Tell me how you did, your learning experience, and if there is something you would change about this learning process. It would be immensely helpful for others starting out with AI and ML. Feel free to shoot me a note and please do give this post a BIG LIKE.

All the best,

Suff

P.S. Connect with me on LinkedIn or follow me on Twitter