AUG 14, 2023
5 Imaginative Data Science Projects That Can Make Your Portfolio Stand Out
OCT 05, 2022
In my past content, I’ve highlighted this important fact: Your portfolio should differentiate you from other candidates. It shouldn’t make you look the same as them. If you all have similar projects with similar methods, how are you going to stand out enough to get hired?
Image by author
What does it mean to stand out though? Standing out can be through novelty, impact, skill, or creativity. In this article, I will highlight 5 of the most interesting data science project ideas, coupled with concrete examples, that will be sure to turn heads.
If you prefer a video format, consider the following to learn more about this topic:
Are you looking for a project that is out of this world… literally? Well, outer space is a good place to start. NASA collects and shares a huge amount of data about its missions, research, and activities for free online.
With this incredible dataset, you can look up data by planet and by mission. You even look into the life-science data that NASA has been collecting from anything ranging from humans to beetles. They even have human-resources data from NASA available to you.
A simple project with this data could be a planet-comparison dashboard. Someone with prior knowledge or interest in space could do something even more impressive, like creating an algorithm to classify space debris or even identify objects in images that astronauts took.
Image from NASA
The next topic is a little closer to home. I have a question for you: Do you consume content? Maybe a podcast, YouTube videos, or tweets? What if I told you that the data from these channels that you love could be the topic of your next project?
One of my own first projects was analyzing the text data from one of my favorite podcasts, Bigger Pockets. Many podcasts publish transcripts of their data for free online because it helps with searchability. So, you can go in and scrape this data and use it in a project.
Image from Bigger Pockets
These are perfect types of projects to explore natural language processing techniques, topic modeling, and some unique styles of visualization.
If an interviewer also likes the podcast you did a project on, you will have plenty to talk about!
It isn’t just podcasts either, in particular, YouTube has a good API for scraping video descriptions, and there are some fairly reliable libraries for getting Twitter data as well.
There aren’t many things that are more valuable than a project that provides real-world impact. Fortunately, there are plenty of free datasets that allow you to explore and potentially influence phenomena in the real world.
The first of these is the data from the Food and Agriculture Organization (FAO) of the United Nations. This is data that I personally explored for the first challenge in Z by HP’s Unlocked movie. That exercise is a great place to get started, but it is only the beginning of the data that is accessible there.
Through the FAO you have global data on food production, food security, trade, forestry, sustainability, employment, and at least a dozen additional fascinating areas.
This data is great for creating informative dashboards or for doing various forms of time series analysis.
If you live in the US, another dataset that may be directly relevant to you and your community is the FBI Crime Data Explorer.
This set has historic and, in some cases, real-time data on different types of crime by state. Again, this would be great for a descriptive analysis or maybe an inferential analysis to determine which crimes have increased rates based on state-specific policies.
Image from the FBI Crime Data Explorer
Data science is moving faster than it ever has before. It seems like GPT-3, a revolutionary algorithm is now old news. From what I heard, there are actually rumblings of GPT-4 coming out in the near future. The awesome thing about this domain is that you can have access to some of the most cutting-edge technology at a fairly low cost.
A project that could open eyes would be leveraging some of the new language models or generative models into a simple website. A huge portion of data science is knowing how to use models after you’ve built them, and this is a great way to showcase those skills. One that is of personal interest to me is DALL-E 2, which I am still on the wait list for. If anyone can get me access I would be eternally grateful!
I personally enjoy seeing how people find unique applications for GANs and reinforcement learning as well. My friend Nick Renotte has an awesome series where he uses reinforcement algorithms to play some of his favorite games from his youth. If you’re working in tech, I expect at least one of your interviewers would be pretty interested in that project.
The last type of project I recommend is probably the most respected but arguably least attractive to most people. That is contributing to open-source libraries. If you are using a library and can think of ways to make it a little better, why not actually do that? My friend Stefanie Molin was recently a bit frustrated with how some of the graphing methods operated in pandas. It was a bit clunky and she thought it could be improved upon. It took her an afternoon, but she made some adjustments and submitted them. Guess what? They were approved!
I know this seems very advanced, but it doesn’t have to be. In a lot of libraries, they are in dire need of code examples of implementation. Sometimes just updating the readme with examples can make you a contributor to various projects. Here are some ideas for open-source data science projects:
This is something that can very much differentiate you from the crowd!
I did my best to try to give you some flexibility within the domains of these projects. If I gave very specific projects, there would likely be a huge influx of them in the market and then no one would stand out. With that being said, after this article, it is likely that the use of these datasets will still increase and you will need to find additional ways to differentiate yourself again. Fortunately, you can get creative and find your own projects. Check out this video on how to come up with data science projects on your own next.
I hope this article gives you some fun ideas about how to separate yourself from the pack and do something useful with your project work. If you enjoyed this article, remember to follow me on Medium for more content like this and sign up for my newsletter to get weekly updates on my content creation and on additional learning resources in the data science industry!
Until next time, good luck on your data science journey!