Data Science and the news around it have made it a hot field with plenty of job openings, specializations, and high salaries. However, many people don’t understand the critical traits necessary for any good Data Scientist.
5 Questions to ask yourself before Transitioning to Data Science
Am I comfortable being the squeaky wheel? The vast majority of data science work requires us to be critical of the data, motivations, and what our company is optimizing for (growth, revenue, cutting-edge products). We are in positions where we have to tell stakeholders (management and investors) why an idea is biased, illogical, or unsupported by the data we have access to. If you’re afraid of saying no to CEOs (a lot) then work on your data communication skills. Understand you’re not saying no to a project or analysis for no reason, but because for just reasons that uphold our scientific standards. We need to call out bad data, bad practices, and products that can be risky to consumers.
Can I overcome the worst data storage issues? Most companies hiring Data Scientists have vasts amounts of data they’re hiring you to deal with. It’s important to know most organizations haven’t solved their data storage issues. In industry, I’ve seen data sent over in screenshots, horribly mismanaged databases, and data inconsistencies that would shock you. I didn’t expect this when I started working for large companies, but most orgs that hire Data Scientists have had a pretty bad history prioritizing their data management and not just what insights they can gain from them.
Would I rather be in software engineering / How much do I care about the scientific aspect? Data Science processes are NOT just like software engineering, but with more math. While many tools overlap like version control and modularizing code, most of our work resembles those in research and academia than the typical dev shop. Figure out if you’d rather spend most of your time building product features or analyzing swaths of data.
How willing and I to learn detailed statistics? When we boil it down, the mathematical foundation most referenced and used daily is statistics. When I first got started I was intimidated by how much I didn’t know
Am I comfortable with uncertainty? Not all projects in Data Science pan out. Often we’ll sink days of work away to find that our model perpetuates bias or doesn’t solve what we were aiming to solve. Are you okay with analysis passed upwards to management that gets ignored and they do what they were going to do anyway? Sometimes that’s what we deal with. It’s not as cut and dry as scoping out a project and building it.
Data Science Roadmap🗺
I love this article by Kate Marie Lewis on how she built a Data Science career in 6 months without any prior coding knowledge. I hope this inspires and motivates you. Her article shares tools she used to develop different DS skills.
I’m nearly finished with this book, but let me tell you, it clearly oulines the steps, skills, and kind if interview questions you’ll get when trying to develop your Data Science Career. Get 40% off with the code: nlddigest20
My Favorite Things 🌟
One of my new favorite podcasts is the Radical AI Podcast. This mile-high based podcast covers marginalized voices in AI. Their most recent episode is a great one, they talk with Karen Hao, the artificial intelligence reporter for MIT Technology Review.
This is a new pickup for me, but I’m a big fan so far. Race After Technology by Ruha Benjamin is a phenomenal read. It highlights the pervasive bias that lies behind many prevalent algorithms.
I really love this talk, A Gentle Introduction to Data Science, from the 2017 Europython Conference. I love the breadth of content covered and it’s a wonderful talk to watch when you have a spare half hour.
Dataiku released a wonderful guide to learning Data Science from Home. I love seeing this as a calendar, but you can start any time you want and it offers a variety of tasks like spending time cleaning up your Github and data crossword puzzles.
Making an online Data Science Portfolio shouldn’t be difficult, but often it’s harder to figure out where to start. Check out this video by Ken Jee on constructing a Github Pages portfolio in no time.
New Date! Catch my Breaking into Data Science Webinar on Sunday, June 7th.
We’re right around the corner from Women in Data Science San Diego (aka Remote). We have a new registeration form, so make sure you register to get the webinar link!
Follow me on Twitter and RT this tweet from now until June 3rd and I’m giving away 2 Python Humble Bundles. You have a week from today to RT & I’ll announce the randomly chosen winners on June 4th.