Artificial Intelligence(AI) and Machine Learning(ML) explained for everybody

What is machine learning? That's the first question in everything is new and the answer is exactly in the question: our brain needs to define, to classify, to compare and to anticipate the things around us in order to have a faster thinking. Machine learning is part of artificial intelligence and it can be translate as “learn how to learn” a computer or a device.

I really like to compare machine learning with human thinking because I believe there's a lot of similarities. Once we see a thing/situation constantly, it remains in our memory and we can easily operate with it without trying to identify it again.

For example, if we see a photo with the Tour Eiffel, we can easily say what it is, because all the media presented it many times and we already have it in mind(unless other touristic points that probably we saw it once and it's hard to recognize them). So it is with the machines: it can be made to recognize the Tour Eiffel, by giving it resources(photos) and then the data should find common characteristics to classify correct the object. Next time, it will be more easy to be recognized and this type of learning it is called Supervised learning.

Like in our life, sometimes we receive rewards or penalties, if we're doing something good, or in other case wrong. The system can try many variants and then it can check what feedback it would get. This one is called Reinforcement Learning(RL) and is not far from behaviorist psychology.

 To me personally, it was easy to understand a programming language and the machine learning by compare it with human mind. We don't represent a tree each time we are talking about it, that simple 4 letters are enough to figure out what is about, so the language is itself a form of abstractization. Programming is just data and its interaction, like in the real life.

My favorite tutorial series of ML belongs to Google Developers channel and it can be found here:

Popular posts from this blog