What's holding you back?
Why haven't you begun your data science journey yet?
I know what you’re thinking: I thought I signed up for a data science newsletter, what’s this motivational crap.
But the thing is you NEED to identify why you haven’t begun learning/given up in between something this fascinating.
Worry not, I’m here to help!
I’ll talk about some barriers you may have faced along the way and help you get over them with today’s mail.
TYPE I: I need to know XYZ to get started
What’s XYZ?
Python, R, SQL, Excel, Powerpoint, Notepad blah blah blah
Probability, Statistics, Linear Algebra, Graph Theory blah blah blah
Batch Normalization for Meta-Learning (said no one ever, but it’s just as ridiculous as the ones above)
Sure, knowing this stuff helps, but you don’t NEED to know all this.
All you need to get started is a computer (and English, otherwise you aren’t understanding a word of all this).
Once you get started in the learning process, you may need to pick a thing or two from statistics or brush up linear algebra (Don’t worry, I will cover all this stuff too), but you don’t need a Ph.D. in Mathematics.
P.S. If you want a headstart, Statistics Made Easy is a great place to start!
TYPE II: I did get started, but DEF
What’s DEF?
I didn’t understand anything!
I got overwhelmed
I’m not Andrew N.G. (If you don’t know him, don’t worry about it)
For starters, pat yourself on the back. You took the first step!
But that’s when you realized that this is one twisted staircase, full of missing steps and creaky wood with no railings.
This is where I come in. Hop in the lift ;)
TYPE III: I can’t learn right now, because ABC
What’s ABC?
I’m super swamped right now
My laptop can’t handle the load of data science
I suck at coding
I’m not going to lie to you: It does take time and effort to get good at data science.
But Rome wasn’t built in a day!
You can make solid progress with even ten minutes per day.
Also, want to hear something interesting? One of the main reasons I got into data science was because I absolutely hate software engineering.
Want to hear something more interesting? You can execute machine learning codes on the Internet (via Google Colab)
Want to hear something even more interesting? A single human brain generates more electrical impulses in a day than all the telephones of the world combined.
Irrelevant, but pretty cool, eh?
So, which type do you fall under? Let me know!
With your fear and apprehensions out of the way, we can start learning :)
The next post, delivered in your inbox this coming Monday, will take you through the process of a typical data science project. In fancy terms, we’ll be learning the OSEMN framework.
See you then!



When do you suggest one should start working with Pipelines in ML problems?
Does there have to be a significant amount of experience in hard-coding the cross validation and pre processing steps before one moves on to using Pipelines?