MOOCs for Machine Learning
There are many-many MOOCs online related to Machine Learning. Be it stats or machine learning models or even core math, there is a lot of content available online and for someone who is just starting with machine learning on their own, it can become pretty daunting, like it became for me.
Its been 3–4 years since I started into ML as an student and I have gone through a lot of lectures online and in class to learn. Through this blog post, I want to help people who want to learn machine learning online.
I have had the chance to interview with certain machine learning giants recently and one thing that I have learnt — You don’t need to know everything. You just need to know basics and some advance stuff, but you need to know it very well. Keeping that in mind, I have prepared a list of machine learning MOOCs from Coursera, EdX and Udacity. I have also considered a level ‘O’ for people who don’t know programming.
Level ‘O’ : You are new to programming
If you are new to programming, without a doubt, head to CS50 by Harvard, there is no better course than this for somebody who is starting coding. It is the perfect introduction to programming.
Level 1 : Math 101
So, here I know that ‘you know coding’. The first step is to learn some math. Even if you know it, you should go through these courses to brush up on everything — very very important.
- Probability and Data: This is a short course that will build up sufficient (but not necessary) background of probability.
- Stats202(Statistics 202: Statistical Aspects of Data Mining): Although, this course says “Data Mining”, the content is applicable pretty much everywhere in AI — NLP, ML and even Information Retrieval.
Level 2: ML 101
So, now I know that ‘you know some math’. Let’s take the baby steps to ML. A lot of people out there prefer to start with Prof. Andrew Ng’s coursera course but in my personal opinion, I believe its a little too much for beginners, specially with the Octave/Matlab requirement for problem sets.
Keeping that in mind, I will suggest that you start with the following course:
ML Foundations with Case Studies: It’s one of best starters when it comes to starting with Machine Learning.
Now depending upon your interest, you can take any/all/some of the following courses (will add more to this list as and when needed):
- NLP by University of Michigan
- Text mining and Analytics
- Neural Network by God Father of DL
- Regression, Classification and Clustering in detail [HIGHLY RECOMMENDED for anyone who is serious about learning ML)
Level 3: Machine Learning
So by now you know some introduction to ML. It’s not yet sufficient to prepare you for what’s out there. This is where, I would suggest you to skim-through Prof. Andrew Ng’s coursera course, probably without programming assignments.
I strongly advice you to take the CalTech’s course on Machine Learning. It’s one of the most amazing courses. It will push you and pull you to learn more. It has math, it has theory and it has practicality and the best part is that course is the same as it runs in CalTech.
Level 4 : Advanced Machine Learning
So by now, you know a lot in Machine Learning. Where to go next? One way is to go to Coursera and take those 4–5 weeks courses on individual courses on models such as that on Regression, Classification, etc.
Another way is to learn deep learning — probably the biggest buzz word of this decade yet. No better place to start than this : http://cs231n.github.io/.
It’s not too much but it’s good enough to make you learn enough for you to start exploring.
ADVICE: While you are learning all this, please make sure the coding you do and the content you learn is visible to people outside (do check the course/platform policy on this though.)
That’s all folks. Learn it all away.
If you would like to know more about me, please check my LinkedIn profile.
Thank you Flipboard.com for featuring this article.
