A Resemblance between Human Learning and Machine Learning

While I was learning Learning How to Learn: Powerful mental tools to help you master tough subjects at Coursera, and having a background in Computer Science, I started joining dots with Machine Learning and found a remarkable similarity between the two.

But it is not at all a coincidence since firstly Humans created the ML, and then the techniques used by humans to learn is highly effective. It is a foolproof solution and highly tested.

So I am going to write about some of the similarities between both fields -

1. Supervised Learning

How a child learns that a cat is a cat. He is shown multiple instances of a cat with telling him that this is a cat. Thereby he has associated a mental picture of a cat with the word cat. This is exactly how machines learn in supervised learning, where a machine is shown a lot of instances of cat and then after enough training, machine knows how a cat looks like and given a test data it is able to recognize a cat.

This is a cat

2. Unsupervised Learning

Given a set of balls and bats to a child, he will be able to distinguish between the two without knowing the purpose of the two. This is similar to how machines do unsupervised learning. Given a data it will segregate the data according to similar features without knowing what the data corresponds to.

3. Transfer Learning

Transfer Learning in Humans is defined as applying the knowledge in one field to some other related field. This is what exactly a computer also does, It transfers the knowledge gained from one learning process to other related fields.

Ex — Artistic Style Transfer

These Machine Learning concepts have a direct resemblance to the concepts I learned during my course. I recommend everyone to take this course as it will help you a lot in any field you are.