A Beginner’s Guide to Machine Learning –Let’s teach the computers

4 min readMar 24, 2024

Before we jump, into this world. Let’s think:-

What is Human Learning ?

Brain learning, is the process by which individuals acquire new knowledge, skills, behaviours, or attitudes through experiences, ( mistakes we make in past ) study, practice,

and exposure to information.

Simple & Easy Peasy — No Trick here.

But now think about Speed & Adaptability to new changes our mind can have.

“ When new things happened in our life the mind takes time and sometimes it’s difficult to adapt the new changes to our life especially when it’s something we don’t expect such as any disease, Job Loss, or any other changes which we are not prepared for.

Although with time everything became Hakuna Matata one day “

Lastly think about the information we can store in our brains.

That’s where Machine learning is super powerful in remembering the things and storing the data and adaptability to new changes and that’s why it is called Machine learning because machines do not have emotions.

Let’s go:-

What is Machine Learning?

Its — “ Teaching computers ” from patterns observed from data. — Simple is that.

Machine Learning & Human learning is having one thing in common they both get matured ( accurate ) with more life experiences or historical data.

Shakespeare said..

“ Wisdom comes with age & experiences in Human”

In simple terms, more the life experiences or examples we have the more we are supposed to be wise.

In case of machines, wisdom ( Accuracy ) will come with more historical data we expose to our algorithms.

“ In Data Science world, machines decision making capability become more robust with the data volume ”

Let’s understand this with small example:-

Surely we have as many applications as you can think, but let me drill down in simpler words.

See the below dataset :-

We have the Name, Age, Experience, Marital Status, Salary columns in our data set.

The first step is to define the Problem Statement.

The mantra is:-

We define the Salary column as Dependent Variable or Interested Variable which we want to predict based on the Independent variables / Feature Variables / Input Variables i.e Name, Age, Experience and Marital Status columns.

There will be many more in real time for sure..

Let’s Imagine now:-

We have a new employee “Eric”, whose dependent columns are below:-

A task is simple just Predict the salary, based on the pattern learn from John, Nathan, Garima , Alice, Mark and Saurabh data.

It’s as simple

Yes It’s as simple as it sounds.

But , we will see in upcoming time how much effort as Data Scientist goes in to predicting the salary of just Eric.

That’s what we called Regression algorithms. (Predicting a continuous value)

Let’s see with another example,

Imagine I have bank data for Loans and who defaulted.

We have again same Name, Age, Experience, Marital Status, Salary and Default ( extra ) columns in our data set.

Let’s define the Problem Statement.

The mantra is:-

We define the Default column as Dependent Variable or Interested Variable which we want to predict based on the Independent variables / Feature Variables / Input Variables i.e Name, Age, Experience , Salary and Marital Status columns.

Lets Imagine now:-

We have a “Eric”, who want the Loan and his data is below:-

As a bank you want to ensure that he does not default you. Right !!

Can you predict as Data Scientist if Eric will default or not based on his data and pattern learn from John, Nathan, Garima , Alice, Mark and Saurabh data.