5 tips to learn machine learning for developers and beginners

We live in a world in which AI, artificial intelligence, is constantly in the media to
describe the future technologies that will affect our everyday life. Machine learning algorithms are most of the time at the core of an AI system and many developers and computer science practitioners want to understand this subject.
Clearly, to learn machine learning algorithms and understand how to use them, many resources are available.
Just to cite a few:
-Intro to Machine Learning by A. Ng, video course (available at coursera http://bit.ly/1IXp8Lg and youtube)
-Machine Learning by J. W. Paisley, video course (available at edX http://bit.ly/2s3pjmG)
-Machine Learning A-Z™ by K. Eremenko, video course on Udemy (http://bit.ly/2s33KTx)
-Machine Learning for Hackers by D. Conway and J. M. White, book R language (http://amzn.to/2t2Plv2)
-Machine Learning from for the Web by A. Isoni, book Python language (http://amzn.to/2sLsgKw)
-Machine Learning with Python by S. Raschka, book Python language (http://amzn.to/2rJTSi3)

All the material listed above will give an excellent grasp on how to use machine learning algorithms in real scenarios and on the theory behind.

I will suggest to dive into the content bearing in mind the following tips:

1. Choose a programming language and stick with it. At the moment (2017), the most used languages among data scientist professionals are R or Python. Both have great machine learning libraries, up to you.

2. Don't follow the hype, understand the basics. Machine learning is a rich field and it is tempting to jump directly to 'deep learning' or 'neuron networks'. However, you need to understand the basics to make sense of these advanced topics.

3. Focus on understanding the process not the specific algorithm. The number of details of each algorithm can be overwhelming but the essential part is to understand flow: how to go from taking a rough dataset to making predictions using different algorithms.

4. Spend 60% of the time on code, 40% on theory. Understanding the theory is important but put it in practice yourself is even more important.

5. Be aware of different prospectives. There are often different ways to see the same data and, as a consequence, the algorithms. Find different explanations for the same algorithm and try to find different interpretations for the data structure.