Basics of Machine Learning

In these days where the world is interconnected, maybe you hear some terms like Artificial Intelligence Machine Learning or Deep Learning, and you think that these words sound like magic that computers do, and it is difficult to understand, but if you follow whit me, I'm going to explain to you in a basic way how this technology work, why is so popular and where is going to be used.

in this video, you can find a brief introduction stay with me and let's bring some clarity about how a computer does basic things.

computers, weird boxes that do magic things

At some point, maybe you used a computer and a computer is a device for working with information also is an enhanced version of a calculator with little extra functionalities. The information can be numbers, words, pictures, movies, or sounds. Computer information is also called data. Computers can process huge amounts of data very quickly. They also store and display data.

imagine that you have a lot of recipes writing in a book, and you continue adding new recipes each day, in some time it becomes a lot of information to remember, this is the point when a computer becomes very useful because it can process millions of recipes in a few seconds

that's why computers become so popular and useful in our lives

What is a program

A program is just a set of instructions that the computer executes, like a sequence of instructions that the computer follows.

create a program is just write rules for the computer, using some kind of computer language, this language is not very different from a human language, although it has more technical terms.

we give the computer a series of rules, that inside of the program whit a certain process we decide how it is going to be treated for output what we expect.

in other words, when we make a recipe for a cake, we put some ingredients expecting that the cake taste delicious and whit some adjust we can make the cake looks beautiful, a program in computers work the same, whit some rules we can adjust what is going to do on the computer.

How Machine Learning Becomes Popular

as we see, in a typical program we need to make the rules for the computer work sounds good because we control what we want to do inside, but it becomes really hard to manage when the rules are more and more complicated.

imagine a program for calculating numbers from 1 to 10, is something that computers do easily, we only need a few rules for that

now try to imagine how we can make a program for recognizing pictures of cakes, the rules for that are really difficult to traduce for a computer program.

as we can see, some task that humans do easily becomes an impossible task to traduce in the computer language, until 2010, when something called machine learning, the salvation for our problems

Let’s say that you brought a cake.

It was so delicious that maybe you want to make this same cake but unfortunately, you’ve no way of knowing who made this cake so you can’t get a recipe for this cake. All you’ve is some leftover of this cake.

then you try to recreate the same recipe and you follow this process to make a cake.

1. Pre-heat oven to 350 degrees.
2. Grease and flour three 6" X 1 1/2" round cake pans.
3. Mix flour, cocoa powder, baking powder, and baking soda. Set aside.
4. In a large bowl, beat butter, eggs, and vanilla.
5. Gradually add sugar.
6. Beat on medium to high speed for about 3-4 minutes until well mixed
7. Alternately combine in flour mixture and milk to the batter while beating.
8. Continue to beat until batter is smooth.
9. Pour equal amounts of batter into greased and floured round cake pans.
10. Bake 30 to 35 minutes.
11. Check with a toothpick to see if it is done. Bake a few minutes more, if needed.
12. Remove from oven and allow cakes to cool in pans for a few minutes.
13. Place cakes on a wire rack, to them, ow to completely cool.

After you made a cake following this recipe you compare it with the leftover cake and you realized your cake tastes different in so many ways. What you found was that your cake was too sweet, too crispy, too eggy, too burnt, etc.

So you decided to make another cake but this time you slightly tweak the ingredients to make it taste more like the school cake. You reduced the amount of sugar in step 5, added more milk in step 7, reduced the amount of egg in step 4 adjusted the heating time and temperature and so on.

This time the cake tastes more like the cake that you brought but not the same, so you continue the process by tweaking the ingredients.

So essentially what you did is,

1. You made a cake
2. Compared it with the school cake
3. Find the differences
4. Changed the recipe to reduce the difference
5. Made a cake with this tweaked recipe

You repeated this process until the comparison in step 2 was good enough.

The important thing here is that each time you compared the result with the original cake and tried to reduce the difference between them by going back and cook again.

Deep learning is exactly like this, you continue changing the things that affect the program so now is more and more like the one you want. One major difference is that while making a cake you know what process influence which property in what way, like adding more sugar will sweeten the cake, but in deep learning, you don’t know, actually you can't.

Fundamentals of Machine Learning

Typing “what is machine learning?” into a Google search opens up a pandora’s box of forums, academic research, and here-say – and the purpose of this article is to simplify the definition and understanding of machine learning thanks to the direct help from our panel of machine learning researchers.

you can found more information on this site, but Machine Learning is the science of getting computers to learn and act like humans do, and improve their learning over time in an autonomous fashion, by feeding them data and information in the form of observations and real-world interactions.

In the field of machine learning, there are generally three methods that are often employed:

1. Supervised learning
2. Unsupervised learning

There are also methods of learning that are inspired by the brain in the form of computer neural networks. The broad name for this particular method is deep learning.

As a general introduction to the basic concepts of machine learning, I am going to explain briefly.

in this site, you can play whit some machine learning algorithms! (https://playground.tensorflow.org)

There are also methods of learning that are inspired by the brain in the form of computer neural networks. The broad name for this particular method is deep learning.

As a general introduction to the basic concepts of machine learning, I am going to explain briefly.

Supervised Learning

Let’s say you wanted to find a relationship between the number of years someone smoked and their average lifespan. A good way of doing this is to look at a large set of data in hospitals, plot it on a set of axes, fit a data model to it, and then use this model to predict future occurrences.

just like the cake, you have a lot of ingredients for different cakes and then you tried to recreate in a new cake.

This is precisely how supervised learning works at a high level. The user provides the program that is known as a “training set” of data. This data is labeled so that the program knows what the true output/answers are. In the example above, data of the years smoked is provided with the average lifespan so that for each person, the computer knows how long they smoked and how long they lived so it can eventually predict how long a random person will live given that they smoked a certain number of years.

Unsupervised learning

Using the information that is not classified, not labeled and allowing the algorithm to act on that information without guidance.

For example, we have historical data on house size and age and trying to find a way to classify groups of those houses. In this case, we can use unsupervised learning to group sets of houses together based on the inputs of size and age and see if there are any patterns to the outcomes that exist.

in simple terms, machine learning is a way to teach to a program how the data is, and the program "learn" from his own experience, making the program intelligent...

but what is intelligence?.

Artificial Intelligence vs Human Intelligence

Intelligence can be defined as a general mental ability for reasoning, problem-solving, and learning. Because of its general nature, intelligence integrates cognitive functions such as perception, attention, memory, language, or planning. Based on this definition, intelligence can be reliably measured by standardized tests with obtained scores, but why the machines are not intelligent?

they can not solve simple problems, like for example making a cake, if we don't teach them how to start and how the process works and the exact process machine learning is also a core part of AI. Learning without any kind of supervision requires an ability to identify patterns in streams of inputs, whereas learning with adequate supervision.

On the other hand, Human Intelligence is defined as the quality of the mind that is made up of capabilities to learn from experience, adaptation to new situations, handling of abstract ideas and the ability to change his/her environment using the gained knowledge.

as a conclusion, we can think of machine learning as a way to recreate how humans learn in a program that we are going to use in a computer, for doing a simple task that is easy for humans but difficult to translate in clear rule whit the traditional way to make computer programs (algorithms).

References

https://www.youtube.com/watch?v=ukzFI9rgwfU

https://emerj.com/ai-glossary-terms/what-is-machine-learning/

https://playground.tensorflow.org