Machine Learning course by Andrew Ng on Coursera — worth it in 2021?

If you Google the top 10 (paid or free) courses for Machine Learning on the internet, in every list and blog post you will find this course by Andrew Ng on Coursera which is offered by Stanford University. The course has been taken by 4M+ students with a 4. 9 rating. Andrew Ng is considered one of the best teachers in the field of ML. So, is this course and Prof. Ng really worth the hype? Here are my thoughts on this.

Although the course was made in 2011 but it is still relevant today because it doesn’t teach you any framework or skip directly to implement the pre-built algorithm. It covers the theory behind different algorithms, statistics, and other mathematical concepts which are used in Machine learning even today.

The timeline of this certificate is 11 weeks but you can complete the course in 1–2 months depending on your speed and whether you choose to explore any extra topic of the course further.

If you have studied algebra long ago or not at all, don’t worry, this course will teach you most of the linear algebra concepts used. The course provides a broad introduction of a few of the most commonly used models. It covers supervised, unsupervised, and introduction to deep learning.

Apart from all this, there are a few points to consider before doing the course.

Pros

• The course starts from the basics, so even if you wish to change your preferred framework, the concepts behind the topic will be relevant.
• The pace of the course is not too fast or slow and Prof. Ng explained the topics very well.
• The discussion forum has many students in it so there is a chance that you will find an answer for your query.
• The quizzes after every few videos are a good exercise to test your understanding

Cons

• At times, the content of the course is not sufficient to understand the topics explained. You would need to refer other sources to get a clear picture of the topics. This can further extend the time you take to complete the course.
• Few parts of the course do require some calculus but it is not covered, as those topics are optional.
• As stated earlier, this course is 10 years old so you will have to study further in order to find a job or build any real world application.

Mixed Feeling

For a better understanding of the topics, you should complete the exercise. and in case you want the certificate, there is no way out. But the exercises are designed in Matlab, so it becomes necessary to learn them.

If you don’t care about the certificate and don’t want to learn Matlab for some reason the same exercises are available in Python also, but I have no idea about the reliability of this source.

On the plus side you will learn a new language, get a free license for Matlab and Matlab Onramp Certificate.

Overall I would say, that if you are planning to learn Machine Learning, this is one of the best places to start. Many articles that you will read online refer to this course as it is very basic and provides a strong foundation.

Solutions and Notes.

You can download the slides used in the lectures, but I have made some corrections, added a few points and few external links in my slides. The link to those notes and exercise solutions are here.

Certification

The course is free of cost, but if you want to solve the exercises and the certificate you would have to spend 4314 INR or 70 USD. The exercises are an important part of the course and the certificate is totally worth it. There is an option to apply for financial aid, luckily I got it.

What’s Next

Once you are done with this course I would suggest you do some courses on Data Science and Machine learning in Python/R or any other preferred language. If you go with python you will learn NumPy, Pandas, Matplolib, Scikit-Learn etc. but before moving to deep learning frameworks like TensorFlow or PyTorch I would suggest you take this course — Deep Learning Specialization on Coursera by DeepLearning.ai. I have completed this course as well and I will write a blog on it soon.

Thank You and All the best

Please let me know your thoughts and queries in the comment section.

Do give it a few claps if you like it.