Machine learning for everyone who doesn’t know anything about it!
Nowadays all people are in touch with cellphones, tablets, computers and all this “smart” devices every day, that makes our lives easier with all the stuffs that these things offer to us, but these devices are not really the smartest we think they are. These devices just take the orders we gave to them. For example, if we give the order to make a call pushing in the icon of the phone, the cellphone would do that for us, or if we tell Siri or Alexa to put some music, the music would be on, or if we give the order to the lights turn off or turn on, the lights would make it. So as we can see with these simple examples, nothing of these things can be done by themselves, they need and order to be execute.
When we look on the internet to read about machine learning we can find complex or simple articles explaining it. But here we are trying to explaining it in an easier way what machine learning is. Let’s start!
Machine learning is a super trending topic; everybody talks about it but few people know what to do whit it or how it really works.
Let’s go to supposed that Sara wants to buy a car. She is checking on internet and she realizes that a new car costs for example $20.000. She also sees one-year-old car that costs $19.000 and two years old car that costs $18.000, and so on. Sara finds a pattern; the price decrease $1000 each year. That is what we know as regression in machine learning, and is a way of predict some value based on historical data. This example explains on of the basics of machine learning that is predicting results based in historical data.
Machine learning has three basic components
Machine Learning is a current application of Artificial Intelligence based around the idea that we should really just be able to give machines access to data and let them learn for themselves. In fact, the key idea behind Machine Learning is that it is possible to create algorithms that learn from data and make predictions on it.
In order to “educate” the machine, you need these 3 components:
DATA: are individual units of information and used for the purpose of analysis. There are two ways to get data: manual and automatic. Recollecting manual information would give less errors that the other one, but takes much more time, which makes it an expensive way. Recollecting automatic information is cheaper but is not that accurate as manual.
FEATURES: parameters or variables. Factors for a machine to look at. Features are simply variables, observable phenomenon that can be quantified and recorded. For example, an application trying to determine the probability of heart disease in patients. What are some possible features? It could be: gender, age, height, weight, blood pressure, among others.
ALGORITHMS: sequence of instructions used to solve a problem. Any problem can be solved differently. Each method has different precision, performance, and size.
Learning vs intelligence
Artificial Intelligence(AI) systems are machines that think and act as humans do. They have an intelligent brain, just like humans do and Machine Learning is a technique, approach, or process for implementing Artificial Intelligence which involves parsing massive amounts of data, learning from that data, and making predictions based on that.
“Artificial intelligence is a technology using which can create intelligent systems that can simulate human intelligence”
“Machine Learning is a subfield of artificial intelligence, which enables machines to learn from past data or experiences without being explicitly programed”
Quotes taken from: javatpoint
Let’s go to define each of these terms in simple words.
Ø Artificial Intelligence: it is the broader concept of machines being able to carry out tasks in a way that we would consider “smart”. AI is the name of a whole knowledge field, just as math, biology, chemistry, physics. So as you can see in the image is the big circle around the others.
Makes it possible for machines to learn from experience, adjust to new inputs and perform human-like tasks. For example, such machines can move and manipulate objects, recognize if someone is moving, or solve other complex problems.
Ø Machine learning: it is one of the areas that compound AI. It is a subset of AI, necessary for machines to learn from data and take advantage of the patterns it discovers through them to make decisions.
Ø Deep learning: it is a method to use neural networks. tries to emulate the human reasoning process, every time we receive new information, the brain tries to compare it to a known element before making sense of it, which is the same concept that deep learning algorithms use, through a model called neural networks.
We can divide machine learning in two different class
In a simple way:
let’s going to talk about each one:
Supervised learning: let’s take the example you want a machine that predicts how long you will take to drive from home to your workplace. You have to train that machine and you need to create some label data that includes for example weather conditions, time of the day, and holidays. This data is going to be the input. The output would be the time it took you to drive home in some day. You have clear that when it is raining you will take more time to arrive home, and the more it rains, the more time you have to spend in the car to arrive home. So there is a direct relationship between the amount of rain and the time it takes you to arrive home. Also you know that the closer it is to 6 pm, the longer time it takes for you to get home. Your machine may find some relation with the label data. And this is the start for your data model, beginning with the impacts that the rain does when people is driving. And also see that in a particular hour of the day, people is traveling more from their works to their homes. So we can translate this by saying supervised learning is a training that consist of inputs and outputs paired correctly.
The regression subcategory in simple words take a group of random variables to predict another variable and tries to find a mathematical relationship between them. For example, in financial markets regression is often used to find how many and which factors such as the price of a commodity, interest rates, industries, and some sectors influence the price movement of an asset.
The classification subcategory is the process of predicting the class of given data points. An example of this could be mapping a picture of someone to a male or female classification.
Unsupervised learning: let’s take the case of a baby that recognizes the pet of his house. The pet is a dog. The baby recognizes and identifies this dog whenever he sees it. Weeks later, some friends of the family visit the house and bring a dog. The dog tries to play with the baby, but the baby has never seen this dog before. Even though the baby recognizes many features that are similar in dogs (2 ears, eyes, walking on 4 legs, barking, etc). This is unsupervised learning, where you learn from data (in this case data about a dog) but you are not taught. If this would be supervised learning the friends of the family would tell the baby, that the pet they have it was a dog. So we can translate this by saying that is a way to learn by examples.
Now we are going to look at some of the types of unsupervised machine learning.
Clustering: it points to find a pattern in a collection of categorized data.
Association: is a method used to find the relationship between variables in the data and determines the set of items that occur together.
Usages of Machine Learning.
In a summary, machine learning is a method used by computers to learn from data and calculations by using mathematics, and have some applications as follows:
Like Google and other websites analyze the time we spend looking for something and collect all the data to later provide better results.
This helps the companies to take better decisions analyzing all the data that they have to take better opportunities, to see fails, to improve sales, among others.
Every day new viruses are born or existing versions are modified, so that antivirus has an updated database is not enough.
For this, they have automatic learning that is based on the behavior of the malware to detect and stop them.
Prediction in preferences
In websites like Google, Amazon or Mercado Libre, even social media like Facebook or Instagram you can notice that these platforms suggest things that you have been looking before. For example, in the case you had look for a computer, they start to recommend places to buy a good computer with good prices based on what you looked before, and this is due the algorithms of machine learning they apply.
They are based on the searches and things that you have bought or downloaded to be able to know what they should offer you.
The automatic learning algorithms are able to detect and prevent diseases such as breast cancer with even one year in advance, directly influencing the possible treatment and the success rate of healing.
Machine Learning is a topic that is in constantly growing those days, and this is article is a little introduction of what Machine Learning is, its classes and how it is used in real life. Thanks for reading!
Machine Learning for Everyone
Machine Learning for Everyone In simple words. With real-world examples. Yes, again This is Billy. Billy wants to buy a…
A Non-Technical Introduction to Machine Learning
Machine learning is a field that threatens to both augment and undermine exactly what it means to be human, and it's…
ML vs Real-World Problems
In the last few years, the explosion in the field of ML and AI have been significant for many startups to disrupt the…
Important Difference between Machine Learning vs Artificial Intelligence vs Deep Learning
Artificial Intelligence is the technique of developing machines that imitate human behaviour. It provides objects…
Supervised vs Unsupervised Learning: Key Differences
In Supervised learning, you train the machine using data which is well "labeled." It means some data is already tagged…