Data Science learning resources — For beginners

Published in

Analytics Vidhya

4 min readApr 3, 2021

A semi-structured musing on the first 12 months of my Data Science learning journey — with some course recommendations.

About 12 months ago I decided to try my hand at Python — initially I just wanted to learn how to code, and I’d heard that Python was a great place to start for someone who hadn’t touched a programming language since high school.

I work a lot with data and having just improved on my SQL skills to help with my work as a Fraud fighter, imagine my surprise when I found out that Python could take this quest for data fluency to a whole new level of usefulness. Thus began my obsession with online courses.

Where to start?

I dove head-first into a bunch of Udemy Python courses that focused on data science, and don’t get me wrong, some of them are fantastic (and will be mentioned below), but they do tend to gloss over some of the absolute basics. Sure, you can get through by copying what the instructor is doing, but do you really get what’s going on?

Why is this guy using square brackets here and curly brackets there? What’s an iterator? What?!

1. Introduction to Python Training

As a complete beginner to Python and programming in general, this really concreted a lot of the absolute basics for me. This is a general Python intro (not focussed on Data Science) and it assumes you know nothing about programming, which was perfect for me. The instructor is clear and thorough, and a day later my understanding of the basics of Python were locked in.

Whether it’s this course or any of the multitude out there, I highly recommend wrapping your head around the basics of the language before trying to do anything with it.

2. Data Manipulation in Python: A Pandas Crash Course

Samuel Hinton is an Australian Astrophysicist, programming polyglot, certified genius, and is as weird (in a good way) as you’d expect someone like that to be — heaps.

There are tons of Pandas course you should take in addition to this one (like Jose Portilla’s Python Data Science courses) but this particular crash course really helped me lock in a few concepts I was struggling to commit to memory. It’s also part of the included content that comes with a SuperDataScience membership, which has a huge range of courses for a very reasonable yearly price.

If you’re like me, somewhere along the line you’ll find yourself in your 20th mini-course and realise that while you’re nailing the course content, you want to tackle some real-world problems, and all of a sudden you’re tumbling down the slope of Mt. Stupid;

Dunning-Kruger Curve. Img source Marketcalls

3. Udacity Data Analyst Nanodegree

Udacity Nanodegrees are a bit pricey, and their model is different to many MOOCs, but if I’m honest the cost and time limitations really helped my motivation (not to mention that they are constantly going on sale for up to 75% off). They were the first I’d encountered with actual projects to complete on your own, and while it’s not the kind you’d get at a university, you do have a real human to give you feedback as well as a pretty lively bunch of ‘mentors’ in the Q&A section, and a very active student community.

The Data Analyst Nanodegree in particular was extremely useful in pushing me to the point where I could use Python in my day-to-day work — well structured courses with datasets that start to approach what you’d find in the real world (instead of dataframes stacked with numpy.random.randint values — ugh).

In some of the more analytically focused courses you’ll find very small sections on Machine Learning. I wanted to know more, and you probably do to.

4. Machine Learning A-Z™: Hands-On Python & R In Data Science

This.Course.Rules.

Don’t get me wrong, you won’t be a Machine Learning Engineer after completing this one, but this is a seriously fantastic course for a first date with ML. Kirill Eremenko and Hadelin De Ponteves are easily the best MOOC instructors I’ve encountered, and this course is packed with content (for Python and R) about loads of the most commonly used ML algorithms, as well as templates for data pre-processing and the models themselves.

They also include super handy ‘intuition’ sections that explain the concepts of each algorithm without getting bogged down in maths and technicalities.

To be continued:

This post could go on forever, so I’ll wrap it up for now with a list below of some of the many websites, people and sources I’ve found useful or inspirational in one way or another. Sorry to anyone I’ve missed!

I’m still learning, and one of the most powerful lessons I’ve learned is that everyone, even the masters are still learning, and won’t ever stop.

Resources:

Legends to follow on LinkedIn: