Machine Learning: The Intro

Imagine that you need to teach a machine to develop a certain type of skill through data. When we say learn, in this context we mean identifying complex patterns in millions of data.

Machine Learning It is about creating programs capable of generalizing and predicting behaviors based on supplied information. This means that systems are autonomously improved over time, without human intervention. The main objective of Machine Learning is to address and solve practical problems.

Why is so important?

Take an example:
Pretend that companies have millions of data that are increasing exponentially every day, so extracting valuable information from them is a competitive advantage that cannot be underestimated.

Deconstruction of Machine Learning

The Algorithms

As you seen in the image, there are 9 differents images and the same TV Show, maybe if you love a TV Show a little bit darker in your homepage appers and image a little bit scary or dark, but if you love a familiar TV show, you find a image more familiar like the images with the kids and theirs friends.

¿What means this?

That means they are using Machine Learning to personalizate recomendations taking our data and preferences to match with us.

This is a recommendation algorithm based on the content they have, on the user’s interaction with that content and on the activity of other users with similar tastes and preferences.

Machine learning is capable of assimilating a wide range of data (big data), but it does not perceive it as data, but as a huge list of practical examples.

We could say that their algorithms are mainly divided into three major categories: supervised learning, unsupervised learning, and reinforcement learning.

Supervised Learning

Supervised learning It depends on previously labeled data, such as whether a computer was able to distinguish images of cars from those of airplanes. For this, it is normal for these labels or labels to be placed by human beings to ensure the effectiveness and quality of the data.

This means that they are problems that we have already solved, but that will continue to arise in the future. The idea is that computers learn from a multitude of examples, and from there they can do the rest of the necessary calculations so that we do not have to re-enter any information.

Unsupervised Learning

Unsupervised learning In this category, what happens is that the algorithm is stripped of any label, so that it does not have any prior indication. Instead, it is provided with an enormous amount of data with the characteristics of an object (aspects or parts that make up an airplane or a car) so that it can determine what it is, from the information collected.

Reinforced Learning

In this particular case, the basis of learning is reinforcement. The machine is capable of learning through trial and error in a number of different situations.

Although you know the results from the beginning, you do not know what are the best decisions to get them. What happens is that the algorithm progressively associates the success patterns, to repeat them over and over again until they are perfected and become infallible.