With the increasing demand for Data Science, there are a lot of Project Works done in Data Science especially by the newcomers and enthusiasts. The plethora of projects found online contributes to the quantity but necessarily not quality. A good quality project / publication will have a greater impact on your DS Journey rather than a million mediocre ones. In this article, I will compile the tips and feedback of various data science professionals on implementing a quality data science project.
Firstly, keep it meaningful
The first and foremost tip before starting any analytics project is to have a clear aim and requirements of what is to be solved. Many a times, projects end up with generating a dashboard or an ML model without having any problem statement or aim. Having clear problem statement and finding the solution is what industries demand.
Avoid MNIST and Titanic
Not that I have anything against them, but please do avoid this. MNIST and Titanic is only good when you start deep diving into data science. After that point, it will not fetch much value since it is way too simple and has not much impact. There are various such projects out there which can be avoided.
It doesn’t need to be that complex
Just like Data Science is not just ML, DS projects necessarily need not be ML. I can even be simple analytics which creates an impact. Majority of the time, projects are judged based on its impact and usability rather than its complexity. So, find a good problem and make up a good solution.
Opt for a full-fledged project
Continuing the talk from first point, complete the project with the implementation. If it is a visualization, make up a presentation and present your findings and conclusions. If it’s an ML, architect a cloud resource setup and host the model. This will make a much better impact than leaving it half done.
Application or Research side?
Job in tech industry can be application or research side. Aligning your projects accordingly can make you more suitable for the role. Solving any problem is what application side projects mainly aim at while innovation and enhancement is what research aims at. Having publications will be a big plus point for research roles.
Get it reviewed
Keep the ego aside and get to know the downfalls of your project. Get it reviewed by an expert and see for improvements. Reviewing opens up a lot of opportunities for improvement and learning. Judging the downfall is the toughest part — Since we came up with the idea and solution, we will hesitate to see the downfalls.
Try moving away from common data repos
To get a better idea of reality, move away from common data repos like GitHub and Kaggle and try scraping or mining your own data. Not always processed data is readily available for projects. Since DS domain is spread over data + analytics, getting an idea on data mining will be a big plus point.
Thank you for reading!
Useful Links:
Find me on LinkedIn: https://www.linkedin.com/in/vishnuu0399
Know more about Me: https://bit.ly/vishnu-u
