# Understanding Machine Learning

## Introduction

As the era of global digitalization is coming closer, systems are getting smarter day by day. Machine Learning is revolutionizing digital culture. We now live in a time where machines are something we need in our day to day lives. They can now teach themselves without human intervention. This self- learning or self-education can produce insights that are helpful in making proper and productive decisions.

Let's dive into machine learning - what it is, how it works, and how will it change the world we live in.

Machine Learning is revolutionizing the way you use the internet and how you search through the search engines. Almost all of us use Facebook nowadays and it is also the best example of Machine Learning and how it transforms content. You knowingly or unknowingly perform activities in your day-to-day life, which use machine learning.

Let us understand what machine learning actually is.

The term machine learning was coined by Arthur Samuel in 1959. He described it as: “The field of study that gives computers the ability to learn without being explicitly programmed”.

Machine learning can be classified into two categories,

## Supervised Learning

It is learning where we are given a set of data, i.e., training data, and we know what our outcome or result should be. You have an input variable, let’s say “X” and an output variable, let’s say “Y”. You can use algorithms to learn the mapping function from input to the output and the intent is to train the function to an extent that whenever we have a new input data “X”, you can predict the value of the output variable “Y”.

I know X & Y make things pretty hard for someone who is not a math geek. Come, let us understand this by a real-life example.

Suppose we have a basket full of assorted fruits. We have different fruits like mango, apple, banana, watermelon, etc. I want to now train a system to identify the fruits in the basket. We will now train the machine with all the different fruits.

We have an apple in our basket so we will teach the machine that if the shape of the object is round and there is a depression at the top and bottom with the color red, then label it as “APPLE”.

Next for a banana, “If the object is a long curving cylinder having the color is yellow and green, label it as – BANANA”.

So, after the training data is given to the system, you bring a new separate fruit, say a banana from the basket, and ask the machine to identify it, since the machine has already learned how to identify it from the previously provided data it will take fruit as an object and then will classify its color and shape and will name it as banana. So the machine every time learns new things from a given set of data and applies that knowledge to a new object presented.

Supervised learning can be used in pattern recognition, spam detection, information extraction, object recognition in computer vision (can be used in robotics) and so on.

## Unsupervised Learning

In unsupervised learning, we cluster the data into knowledge groups without having any external knowledge or label. The only difference between supervised and unsupervised learning is that supervised has access to labels while unsupervised lacks it.

It is where you only have an input data “X” and no corresponding output variable, it is called unsupervised learning because there are no correct answers and there is no teacher, algorithms are left to their own devices to discover and present interesting structure in data.

The best example that I can give you for unsupervised learning is, you must have gone through the crazy sci-fi movies where the robot learns itself by going through thousands of videos and pictures over the internet, or scans a book in a minute and then he has full knowledge of that book.

Nowadays we can buy a robot sweeper at our home --  a small device that cleans off the floor. If it can be improved with an AI using unsupervised machine learning then he can develop a theory that there is dust below the sofa set. He will go there for a week and every day he will find the dust after a week later, he will develop confidence that there is dust below the sofa.

Nobody explicitly tells the robot but he gains the knowledge by practicing the same things and finding this theory about the dust under the sofa correct.

When machines start making decisions on behalf of humans without involving their input and serving them what they actually want then that will be a whole new era and a great time to live in.