The Final Year Project Advice series by the Innovation Garage brings an article for the Computer Science and Engineering students to provide a helping hand with the final year project work.
The most commonly chosen disciplines in Computer Science and Engineering are ML, Distributed Systems, Crypto, Image Processing, Cloud computing, machine learning, data analytics etc.
Selecting a discipline for your project can depend on several factors. Build something that has never been done before or has been done poorly. Your idea need not come from a “reference paper”. You can build something completely new too. Do not take up a very small project like building some sort of chat bot or some social networking site. Take up something big, something that can make an impact. Think more from a research oriented side than a product oriented side. Work towards achieving a proof of concept rather than trying to build a business solution. If your idea is strong, you can always pursue it later on.
Professors are usually more than willing to help with your projects. Ask the faculties themselves. The teachers have a fair idea amongst themselves as to who is good in what. Other than that, you can browse through the teachers’ interests on their profile pages on the department websites and read up some of the research papers to see if your interests match.
As far as grading is concerned, get your work done and get a good grade. Out of 2 or 3 people in your team, the one who interacts more (read every week) with the prof gets the higher grade. So along with your work, make regular interactions with your prof. It need not be during the prescribed hours in the schedule. Drop in anytime you see the professor is in the cabin and ask for an hour to sync up on progress. If the professor is busy, schedule a meeting on the same day or next day.
Execution of the project is key. Don’t finalise the project till you figure out it’s feasibility. After finalising the project, focus on drafting the architecture and the solution to the problem you have chosen. This should be over in first sem. In second semester, you can start building a proof of concept or in some cases a formal mathematical proof. If it involves a lot of code from ground up, you are doing something wrong. You should use libraries wherever possible and write lesser code. again, the idea matters. whether you use camel casing or some other coding convention or how good OOPS you use in your code, it will hardly matter as the code is just a small example of your system working.
Keeping these tips and techniques in mind, the final year project will definitely become much easier to tackle.
Contributor: Saumyajit Dey