The Data Conundrum: Learning Data Science

My journey into the fantastic world to understand data science began with my attraction towards a tool: R. While looking for some insights on Capital Assets Pricing Model, I stumbled upon a blog post which used R to derive, calculate, the CAPM model with images and loading data online, it seemed fascinating. The fascination back then with R, however, could not last very long. One: I wasn’t engaged in any project/or task that would give me the better advantage of using R and most importantly, I was not sure what data science is about. If you notice after 2013 so many online platforms popped up talking about the future of data science and why all should be learning it. Then books after books on big data etc.

My idea was still immature and I tried to explore many resources for many ideas from books to podcasts, the many Youtube videos, and courses offered in Udemy, Edx, SuperDataScience platform.

After 2 years of running here and there into each link and other sources, my conclusion (for now) is:

1. We need to be able to see the big picture. Do not make a mistake of learning any subject for earning a raise or landing on a job.
2. Fall in love with storytelling. I have heard complaints from some top executives that the data science team in their company has been a huge burden (what a CEO wants is a 5 slides on what decision he should be making rather than a data science team throwing a 20 slides of the team’s work and assuming data would be enough to tell the story itself).
3. Try to fall in love with programming. I love Python as a tool and have equal respect for R, Power BI, Tableu, and other forms of languages that help us to tell or foretell a story. I prefer word foretell for machine learning.
4. Statistics for storytelling. Do not worry! Please do not start proving any theorem.
5. Lots of reading, or just find a podcast that keeps you on sight about the goal you wish to reach.
6. Projects and practice. Go to Kaggle!

But who is going to teach you all of these? Sadly, you might have to wonder and track for resources or choose your online or offline mentor. Offline mentors and community would be fantastic! Instead of saying you want to learn to code in Python rather put a narrative of how learning python can enable you to solve big challenges like climate change or automate some mundane tasks.

I regret that I was not into programming (we were forced to learn C and C++ in undergrad, however, we as a student never explored the subject more than from an examination point of view). If you just googled learning about data science or data analytics in 2020 or you think you are too old now to learn, let me tell frankly your timing is perfect.

The importance of data analyzing and learning to visualize the data to tell a compelling story would be in higher and higher demand. Maybe soon an AI-based solution would appear and take this whole role from humans. But for now, if you are into learning data science the above 6 points would take you somewhere ( you will be frustrated for a couple of months ). At least you would be able to know this whole conundrum is worth solving or not. Or, do not worry, if you give up on machine learning or the math behind it, you can switch to areas related to it which uses the tools in data science for fields like digital marketing analytics.