My Blog

Linear Regression

Want to see an example of
how linear regression works?

Think you need to decide a price for your product and you have the knowledge of current market of your product. Lets take the 5 most important things that can set your product market and then we ask you to estimate the price of the product after you having the knowledge of the market and knowing how much each of the 5 factors influence the market. Thats what Linear Regression Model does for you.

This is how a simple Liner Regression Model.

K-means Clustering

An example of K means clustering so
you don't bug me about this in an interview.

Lets suppose you are assigned to lead a huge project and have a huge number of people who work under you, and you know the 5 most important thing in your line of work any worker should have in his resume to know which team under you he/she shall be working with on projects for maximum efficiency. Now someone get a job under you and you need to assign him his group what will you do? You will try to match his traits with people already working there and then try to do your best in placing him in a group he would be most efficient working with, right? Thats what K-means clustering model does for you.

This is a simple KMC model

KNN

The working concept is similar to
that of K-Means Clustering.
Both of different purpose tho.

Okay think you are a leader of a small group in which there are around 150-200 people yea don't get cocky tho, so you have a new participant now and you want to put him where he feels he belongs so you ask him 4-5 questions to judge his personality a little bit. Now you have asked these similar questions to everyone that joined your "small group" so you know what they all responded now you match his answers with everyone and then see which people have the closest answers to that and then you check in which group are the maximum people from and then you put him in that group, thats KNN for you.

Here's the link to a siimple KNN code.

SVM

I would prefer you imagine Pirates of the Caribbean
Dead men tell no tails last scene in which
the ocean separates from between that helped me.

Lets assume you are your class monitor and your teacher tells you to divide the class in 2 teams to play antakshari. So you go ahead divide the class in 2 teams including yourself but then there is one kid thats left now since you have 2 teams completely separated with people who have in common or are friends you have that one kid to assign. So you look in which group he has more friends and the group he has more in common and assign it to that group.That's kind of like what SVM does.

Here's the link to an example of SVM

CNN

Okay think you are playing hide and seek with your friends. Now its your turn to play so you know how your friends look like obviously but your mind isn't really good at finding your friends when you are running and your vision is actually getting a little blurred because of running so fast. So if you had a better idea of how the person looks is he tall or short, is he slim or not that would really help ya if you are just trying to find the same face as your friends in the public to find him. So this is what CNN does and its very beneficial over other algorithms in recognising faces.

Here's the link to an example of CNN

Django Project

Django was a little tough at start but it has a lot
more easy to use functionality once you are familiar with
it. NodeJS is also cool but kind of tired of getting
and pushing requests.

Okay working with Django felt like doing a project but throwing all the relevant notes all around the room and then using the most important ones when you learn how to use them. There are a lot of files to work with and none can work alone you need to learn them all to gain complete functionality.

Here's the link to an example of Django Project