An Introduction to this New Kid Called MACHINE LEARNING

Published Jan 04, 2018

MACHINE LEARNING

That was good enough to grab everyone's attention, I guess.

Before exploring this field any further, I would like to ask you.

What do you think machine learning is?

First things first:

It's Math, Not Magic. PERIOD.

I'm sorry, I had to break it to you.

Let's begin.

Have you played this game?

The objective of this game is to toss that paper ball into the dustbin.

Sounds easy?

If you haven't played this yet (seriously?), or you feel like playing it again, I'd say go for it.

Did you try it?

Attempt 1 — You realize you need to put in more/less force.
Attempt 2 — You realize you might have to adjust your throw angle.
...
Attempt n — Success!

Did you realize what just happened?
With every throw, we learned something new, and we kept learning until we achieved our goal.

Practice makes perfect.

And that is exactly what machine learning is!

When we program a machine to learn from an experience (Collected Data) to improve its output, it's called machine learning

Here's the formal definition

Machine learning is a subfield of artificial intelligence (AI) concerned with algorithms that allow computers to learn. What this means, in most cases, is that an algorithm is given a set of data and infers information about the properties of the data — and that information allows it to make predictions about other data that it might see in the future.

What are the different skills and information do you need to cook something?

• Skills to Cook
• Kitchen
• Recipe
• Ingredients

Now, to program a machine learning model, we need exactly the same things.

• Fundamental Skills : Python, Linear Algebra, Statistics and Probability ~ (Cooking skills)
• Machine learning Setup ~ (Kitchen)
• The Machine learning Design Cycle ~ (Recipe)
• Data ~ (Ingredients)

If you know the RECIPE, then with the right SKILLS and INGREDIENTS, you can easily cook a machine learning model in your KITCHEN.

Now, let's take a look at some interesting applications of machine learning.

DEMYSTIFYING PRISMA: NEURAL ART

You might know about Prisma — it was a thing in 2016. Most people thought that it was just another photo effects app. Well, it was a photo effects app, but on an another level.

How Prisma worked was that it was given some famous painting and their machine learning model tried to understand the painting technique used by the painter. It then applied those techniques on the user's photo.

IMAGE CAPTIONING: NEURAL STORY

Image captioning is the process of adding meaningful descriptions to images (used when either the image fails to load or by assistance software for the visually impaired).

The following images were given to a image captioning machine learning model model called Romantic Novels.

You might argue that these captions aren't accurate or correct. Well, mind you, you've seen the movies or the news, not the machine learning model. Moreover, the machine learning model was trained on fourteen million passages of romantic novels. Read more here.

This was just a glimpse of the beautiful world of machine learning. Do tell me if you want to see more such articles or would like to see explanations of various machine learning algorithms. Cheers!