Machine Learning v/s Newborn Child

Hello World,

In this blog I will explain briefly, the very idea of Machine Learning through a common world example -A Newborn baby. I talked about this in my earlier post too.

My previous blog: What is Machine Learning?

There are many ways one can understand or get a intuition about Machine Learning such as books, videos, or a Professor. But the best way to learn something is through Nature, right ?

So let’s try to understand Machine Learning (supervised machine learning to specific) through Natural Phenomenon.

Now imagine a Newborn child i.e. A child within 4 weeks of his/her birth.

Photo by Adele Morris on Unsplash

This might help. Cute, isn’t it ?

Now when a child is born what does we know i.e. What tasks does he knows how to do when he comes inot the world? What prior knowledge he has ?

These are many questions and the answer is nothing except breathing (but also breathing is involuntary). A newborn does not know how to walk, speak, put there head straight, or to write a blog.

From this we can consider a newborn (putting aside all things that make us humans like emotions and feelings) as a bare machine.

Human vs Machine

As the above picture illustrates Machine/Computer is somewhat similar to a human/baby.

Just like that a baby’s brain when he is born is similar to a empty hard disk and a CPU. But in further years the same baby can speak, walk, run, jump, etc. So how after some time a baby is able to do all these things? The answer lies in one statement which we heard all our lives.

Practice makes a man perfect.

That is correct. A baby observes a task, then practices to repeat the tasks and overtime, He/She learns to perform the task. I took a similar example in my previous post -What is Machine Learning?

We’ve seen if a child is born and brought up in Indian family his first words are in Hindi, if in American family, his first words are in English, if in Spain then Spanish and so on. But this is not taught specifically he/she learn this on his/her on.

This happens because of Language Acquisition- the child hears the voice again and again and thus slowly slowly becomes aware of words a small step at a time and as said “Practice makes a man perfect” implies.

So thats the secret, practicing again and again is how one learns how to do things.

So now we know what the word learning in MACHINE LEARNING implies.

Process in Machine Learning

Although the task of learning in a machine is achieved by the advanced Mathematics. The idea on which machine learning stands lies in process Observe →Practice →Repeat.

Data is provided to an algorithm and it runs complex computational or rather say Mathematical operations on the data to form a relation. The first outcome in general cases is the worst possible. As we keep on feeding more and more data and run the alogorithm again and again (i.e. more practice) it improves itself overtime and after a time it provides good results.

So there you go. Now you know what is Machine Learning and how it related to the real world and Nature.

I really hope you enjoyed reading this post. I really appreciate your responses. If you liked this post or it helped you in anyway. Please Clap👏.

To see my projects: Github

Thank You