Introduction to Artificial Intelligence, Machine Learning, and Deep Learning

Introduction to Artificial Intelligence, Machine Learning, and Deep Learning

Imagine you are living in a smart home. You wake up early in the morning and your personal virtual assistant realizes that you are awake and automatically pulls the curtains on the window for you, following with a weather update. This is all the mechanics of Artificial Intelligence(AI).

So, let’s start with the basics first. 

What is Artificial Intelligence?

The term ‘Artificial Intelligence’ was first coined by John McCarthy in 1955 at Dartmouth Conference. Of course, there was plenty of research done by others on the same subject as the research of Alan Turing, but it was undefined at that time.

Definition of Artificial Intelligence

“Artificial Intelligence is a branch of computer science which is dedicated to building intelligent machines that can recognise human speech, objects, learn things, and solve problems like humans”.

The concept of Artificial intelligence is to build machines that can perform human functionalities like the capability of thinking like humans and mimicking them as well.

Major features of AI include-

  • Speech recognition
  • Object detection
  • Problem-solving and learning from given data 
  • Planning for the future

Types of AI

1. Purely Reactive - It is the most basic form of AI. As the name suggests, it doesn’t store memory or any past experience, it reacts directly to what it observes.

Example: A computer program that plays a game of chess with you is purely reactive; it sees the move of the opponent and counter it in the best possible way.

2. Limited Memory - These machines add up on pieces of past information into preprogrammed representations of the world. It just has enough memory or experiences to make proper decisions and execute appropriate actions at certain events.

Examples: self-driving cars, automated chatbots.

3. Theory of Mind - This type of robot or machine has thoughts, human emotions, and can interact socially. This category is yet to be explored on an advanced level.

4. Self-Aware - These machines are considered the future of our world. These are super intelligent, sentience, and very conscious.

Artificial Intelligence can be achieved in two ways. 

  1. Machine Learning
  2. Deep learning

 

Machine learning

Machine learning is a sub-area or sub-set of Artificial Intelligence. Machine Learning enables computers to act and make data-driven decisions rather than being explicitly programmed in order to carry out a certain task. These programs or algorithms are designed to learn and improve over time when exposed to new data.

Let us understand Machine learning with an example. 

Can you identify these animals?

     

Yes, of course, you can. 

Now, is it possible for a computer to do the same?

Hmm!

Well, it is easy for humans to know the difference between a dog and a cat. But for a computer, it’s not quite possible. When you consider the physical appearance as the difference between cats and dogs you can say cats have pointed ears and dogs have floppy ears. The other differences may be the tail length, fur texture, colour, and so on. This means a lot of factors to program manually to help a computer spot the difference. Machine learning is all about making machines learn just like humans. And like any toddler, they learn by experience.

Other examples of machine learning:

  • The Ola App automatically determines the distance, estimated time for travel, and fair price.

(Source - Ola Cabs)

  • The Netflix app knows what you want based on your search history, so as Facebook feeds.

(Source - Netflix)

All these are possible because of machine learning. 

Deep Learning

Deep learning is the next evolution of machine learning. It’s a subset of machine learning that is inspired by the functionality of our brain’s neuron cells which give birth to another concept called Artificial Neural Networks. This neural network enables machines to analyze, understand, and take decisions on their own. These neural networks mimic the human brain.

 

How Deep Learning Works?

Take the example of a machine that recognizes animals.

Now, let the task of the machine be to recognize whether the above animal is a cat or a dog?

In machine learning, we have to pre-define the feathers of the animals like:

  • Check whether the animal has Pointed Ears or Floppy Ears
  • Check the Tail length
  • Check the animal’s Fur Texture and Like this many other features have to be checked

We will define the facial and physical features of animals and let the machine identify which features are more important in classifying a particular animal. 

But deep learning takes one step further, it automatically identifies the features which are most important in classifying animals. Deep learning requires a huge amount of data and a very complex hardware system whereas machine learning can give results even in lesser data and works with simple hardware. 

Thank you for reading this article. Please leave your feedback in the comment section.

Written By



Karthik KS

Karthik KS works as a Senior Digital Marketing Executive and Technical Content Writer for Skill At Will. He enjoys writing about different technologies and career trends. He is an expert in Search Engine Optimization, Social Media Marketing, Google Ads, Google Analytics and Google Tag Manager. He has done Masters in Marketing and PG diploma in Digital Marketing.

Shares

No Comments

Write Your Comments

Submit

Related Blogs

Related Courses