An Initiation To AI, Machine Learning and Deep Learning For Everyone

In this blog, I attempt to explain what AI ( Artificial Intelligence ) is, and it’s relation with Machine Learning and Deep Learning. This article is not for technical people only. I write it for all those who are inquisitive about these emerging trends.

We all are well aware of what intelligence is. The ability to learn, and to use this knowledge to do something useful, can be interpreted as intelligence. This implicitly applies to living creatures, humans for us. When machines/systems take decisions and perform informed actions, it is called Artificial Intelligence. Simple this far.

The question that arises now is, how do we infuse intelligence in systems? One way to achieve this is to write software program with rules and guidelines, telling the machine to take specific actions on certain triggers.

Another way to realize this goal is, to feed huge amount of data, aka information, to the system. The system learns to discriminate/distinguish on the basis of what it learns from the data. In very raw terms, we can say that when the system learns and trains itself from data, it is called Machine Learning. Machine learning is a way of attaining AI.

Let me explain this with an example. If we want the system to tell if the image is that of a rose flower, we feed it with huge number of images of flowers that have been tagged by users as rose. This means that the images have information that the flower is a rose or not. The algorithms then create a prototype that can tag images that contain rose images,versus those that don’t.This way, our system knows when it sees a rose, and is able to distinguish one on seeing it, just like humans.

Deep learning is a subset of Machine Learning. Artificial Neural Networks (ANN) are the building blocks of Deep Learning systems. ANN emulate the human brain and its inter woven network. It consist of neurons like human brain. They are made of numerous hidden layers, in between input and output layers. Each hidden layer recognises/learns about a particular aspect or feature. In the rose example, there may be one layer that recognises the petal shape.

As we have understood by now, AI and Machine learning involve analysis of large amount of data.This data is used to make predictions for the future or the task at hand. For example, if we need to predict a customer’s buying preference, we analyse his previous purchasing habits. On the basis of this, we predict his choice for the next buy !

Python and R are the common software platforms used to support the work in AI. Both these platforms have numerous packages and libraries that provide algorithms and interfaces for predictions, statistics and Machine Learning.

It is being forecast that many new jobs and avenues will be created in the field of AI, Robotics, and IoT in the coming future. Let us start preparing ourselves for all that is about to unravel in AI ☺️ . I hope this blog got you started in that direction.