How to learn Machine Learning - for beginners

This post helps you to understand if the Machine learning by Andrew Ng course from Coursera is good to start for a newbie.

ML and AI has become a buzz word now a days. The world seems to be moving slowly from Mobile enabled applications to AI enabled applications and so are the developers. Not just that, even the concepts and applications involved in AI making developers enthusiastic and driving them to learn these new topics.

So where to start?

Bigdata, datascience, ML, AI – You might have heard all these are interrelated and requires complete understanding on all to start with ML, but that’s not true. To start with Machine Learning, you do not need any prior understanding/experience of data science/Big data or any other tools but just your passion towards it.

Machine learning is nothing but applying few algorithms for real time problems. So learning the ML involves the below steps

Getting the list of Algorithms

There are many algorithms used in ML. Getting this list can make you understand the boundaries of ML but these boundaries are not definite. As the ML is still evolving, you might see new algorithms evolving. So be open to new algorithms.

Understanding why and when each algorithm is used.

When we say algorithm, it is some defined steps or could be a mathematical equation with some conditions.

Each algorithm would have definite purpose and a history. So understand how that algorithm/equation has evolved. however you need not to remember/memorize the steps to come to that equation. Understanding the concepts should be enough. On the first attempt you need not to do a deep dive. just understand what they initially and based on the problem and need you can have a deep understanding of each of the algorithm.

Applying a proper algorithm to a problem.

Start with simple problems listed on Kaggle.com and other competitions. Try to solve them on your own and then compare your results. This will actually start your journey towards the ML. As you keep solving the problems, you get more acquainted with the algorithms and concepts and eventually you will be able to solve complete problems. This is a continuous journey and the more you do this, the more expertise you gain on ML.

The ‘Machine Learning by Andrew Ng’ course in coursera is a basic course which covers everything we require. So it is good to start with this course and understand the depth and maths involved in ML. It also covers some problems for practice on each topic it covers.

Is this course enough?

No. Never. Unless we try to solve real time problems it is not possible to become a Data Scientist. Kaggle has wide range of problems on data science to start with. Start with some basic problems and then slowly with complex problems.

Suggested learning path

If you are looking for a complete tutorial on Machine Learning then I strongly suggest to go through this course on Machine Learning from Udemy - Complete Machine Learning course for beginners

I suggest this course as it doesn't require any prior knowledge on any of the concepts, easy to understand and a hands on tutorial which easily helps you build a Machine Learning model.

Good luck!!