How to pick up a data science project to work on?

A couple of days ago, a junior brother from my engineering college reached out to me asking if the data science project he had chosen was actually worth spending a year of time on or not. 

Typically, in engineering colleges, you get the opportunity to work on a project of your choice with a group of friends for a whole year but first, you need to get your project approved by some designated teacher. So, this brother of mine was trying to select a project on which he could work with his friends and eventually, build something that he could submit to get good grades.

I obviously gave him some advice on what to do and what not to do and I thought putting up a video about this would be helpful to a lot of people in the same or similar kind of situation.

You may be looking for a data science project to build your portfolio or for acquiring and implementing new learnings, so here are three suggestions I’ve put together to help you pick the right data science project:

1. Pick a project that you like

This may sound ironic but when you actually graduate from college, or to be more specific, when you actually start working at a job, you are often compelled to work on projects that you may not want to work on. So, when you have a chance of picking up a project of your own, it’s a good idea to select the one that you like the most.

Most of the time when you work on something that you like, you learn faster and you start becoming an anticipator rather than a procrastinator. So, when you have a list of 10-20 different data science project ideas off of Google, start by going through all of them and pick out the idea that would make you a happier person by the end of it. If you cannot find such an idea, google some more.

2. Pick a project that helps you learn

In my final year of college, I and my friend went out and found the most exciting and complex data science project we could work on at that time, that is, audio to image generation. The idea was to build software that could take a sentence of speech and convert it into a visual representation. So, if I said, “Black dog inside of a house” the software would synthesize an actual picture of a random black dog inside of a random house.

Again, this was a complex project for us, and the reason we chose it was because we wanted to learn how to do this. By the time presentation day came over, we could only get the software to a point where it could generate pictures of dogs and houses but not as accurate as we had wanted it to. So, did we miss the mark that we had intended? Yes. But, did we also get a ton of knowledge and great grades, definitely yes. 

So, if you like a project and if it helps you learn a lot of new things, you’re getting closer to selecting the project you want to work on. By the way, selecting a project that is out of your comfort zone will also train you to become a quick learner and a brave researcher.

And, here’s the third and final suggestion:

3. Pick a project that helps you stand out

In Machine Learning, you generally start by learning about classifiers and regressors and so do your friends and peers. But, you can easily stand out from the herd by looking at even a simple regression project from a different and unique viewpoint.

Let’s say you’re building a house price predictor. There are several ways to go about this and most people would approach this project by using features such as the floor size or/and the number of bedrooms of the house, etc.

But, what if you could build a house price predictor that predicts the price based on the images of the house itself? This simple shift in viewpoint makes the project exciting, complex and helps you stand out at the same time. This is what I meant by picking a project that helps you stand out.

Conclusion

So, the next time you go looking for a data science project, please remember these three suggestions.:

1. Pick a project that you like
2. Pick a project that helps you learn 
3. Pick a project that helps you stand out

Here’s a video that talks about all the above from our official Youtube channel:

Hope this article was helpful! If you have any thoughts of your own, please feel free to comment it down.

Do you want to learn Python, Data Science, and Machine Learning while getting certified? Here are some best selling Datacamp courses that we recommend you enroll in:

1. Introduction to Python (Free Course) - 1,000,000+ students already enrolled!
2. Introduction to Data Science  in Python- 400,000+ students already enrolled!
3. Introduction to TensorFlow for Deep Learning with Python - 90,000+ students already enrolled!
4. Data Science and Machine Learning Bootcamp with R - 70,000+ students already enrolled!