What is Machine Learning?

In this article we will look at what is machine learning, what is deep learning and some real world applications of machine learning.

 

Source: readspeaker.ai

Introduction

First let’s talk about at what computers are good at and at what humans are good at. Computers can perform large calculations within milliseconds or even in seconds. Computers can search through millions of records (web pages for example) within nano seconds or in a few seconds. Humans take more time to perform the afore mentioned tasks.

 

On the other hand there are some tasks where humans can out perform machines (computers) such as driving a car. Another example is carrying out conversations in natural languages.

 

What is Machine Learning?

 

Machine learning involves making computers learn things like we humans learn. So, how do humans learn to drive cars and speaking in natural language? To understand this we must know how a human brain works.

 

What is Deep Learning?

Human brain consists of billions of tiny structures called neurons. When a human thinks or makes a decision, a chemical signal is generated which lights up these tiny neurons. In the below picture you can see a simplified representation of these neurons.

 

 

The circles are neurons and the arrows are the synapses that connect these neurons. When we are babies and when we see a rabbit for the first time, a specific set of neurons fires up (red arrows in the belowfigure) and the output neurons corresponds to separate entities like a rabbit and a dog.

 

 

When the baby sees another rabbit on the next day or later, the a set of neurons fires up and the baby knows that it is a rabbit. Similarly when the baby sees a car, another set of neurons fires up (blue arrows in the below figure)  and the baby knows it is a car. This process is known as learning/training.

 

 

So from childhood to adulthood, a person learns different things. Different neurons and synapses fire up and communicate with each other to learn and make decisions. Computers can also be trained in the same way humans are trained.

 

We create neural network models and we train them using lots of training data to decide on a corresponding output and these neurons and edges are adjusted to learn so that when we give the same input again, the neural network model it can give the correct output. This is called deep learning.

 

 

Machine learning is a broader term (as shown in figure below) which is a set of techniques to make computers learn and deep learning is one of those techniques. There are other techniques in machine learning which can be used to extract knowledge from available data and use that knowledge to make decisions.

 

 

Machine Learning Applications in Real Life

The spam engine which classifies unwanted or phishing emails as spam in Gmail or Microsoft Outlook is an example of machine learning.

 

Source: blog.comodo.com

 

Personal assistance devices like Amazon Alexa and Google Home are another example where machine learning is used for natural language processing.

 

Source: donsnotes.com

 

Youtube’s recommended videos (recommendation systems in general) is another example where machine learning is used, where they look at your past watch patterns and predict the videos that you might watch.

 

Source: theiconic.tech

 

Driver less cars are another example where machine learning and artificial intelligence are used to detect and recognize objects, decide in which direction to go etc.

 

Source: time.com

 

For more information you can visit the below links:

• What is machine learning? (Youtube)