What was your thesis topic?

Question

IT education

At our department, as always, they spontaneously announced: "in a week we will submit applications with the topics of our bachelor's theses and the signatures of your supervisor (whom you yourself must find)". Nobody was ready for this, because everyone expected that it would be necessary to decide on the topic at the beginning of the next semester (i.e., somewhere in February), but neither now nor in a week ... Therefore everyone, including me, is in a panic ...
In this regard, I would like to hear what topics your diploma projects had? Interested in IT topics related to programming . I myself cannot unequivocally define the area of ​​interest to me: at the university there was a sea of ​​​​theory, little practice, so the only thing that caught on from the less theoretical was neural networks(I wrote a term paper on solving the classification problem with a single-layer perceptron, in principle I liked it, but in terms of literature there is a sea of ​​\u200b\u200btheory, and very few examples of real programs, I even had to write a seemingly very trivial network blindly, without real examples). I am also interested in working with the network (I just work in this area). Thus, I would like to somehow beat these two topics in my project by creating some kind of client-server application using neural networks :) Although, it is possible without the latter, if it is something worthwhile...
Just what do you think? I just can’t think of something that hasn’t been invented yet and that will be really useful in practical use ... For example, at work I ran into ICE (Internet Communication Engine) from Zeroc (simply speaking, a wrapper over tcp, udp, ssl and + a bunch of services for working on the network), a cool thing, extensive documentation, but in the process of use, minuses are also revealed. For example, at a low level in tcp, you can find out about the state of the connection (it is active or broken, and, in the second case, forced or "by all the rules"), but this was not implemented in ICE. Those. there was an idea to write your own data transfer protocol, but this, in fact, again, will be a regular wrapper, and why would anyone need it ...
Ideally, this should be a topic in which some research could be done, i.e. so that there is a lot of theory on this topic, as well as write some useful software product... Languages: C/C++/Qt and asm ix86 at the amateur level :)
Well, in general, okay, something dragged on the story... I wanted I would like to hear your topics (or perhaps advice on this matter)
Thank you all in advance!

Answer 1

As an option: "Classification or clustering of information", for example, there is a certain storage of books (dump), you need to automatically rake the dump of books into categories, ideally, you need the program to determine the categories with subcategories and scatter the texts into folders or at least in the database. You can work not with books, but with a repository of distributions with a brief description, or blogs, or a knowledge base in the form of (questions and answers), it doesn’t matter. Those. to make a certain resource where the information gets after which it is analyzed and classified. In this task, you can intertwine both neural networks and different network technologies.

Answer 2

My thesis is in CFD. A wonderful opportunity to apply the knowledge of mathematics, physics and high performance computing. The problems and technologies are many. To process tasks, MSU provides computing power. Therefore, you can achieve good performance, and such specialties are required by many companies. For example, Apple (calculations of heat distribution), Google (there are very high requirements for the vacancy in terms of experience, heat distribution tasks). Boeing (problems of aerodynamics). Dry Company. Automotive industry, almost all companies. Also F1 teams. The list can be long. Now I work at the Federal Nuclear Center.

Answer 3

I stupidly wrote a system that collects data from different places, puts it in a database, and then reports are built on this database in Excel with graphs. For a long time the brains hovered about the lack of scientific significance, but, as it turned out, it was enough for the commission to show the act of implementation, and, unless some stupidity was blurted out, 5 was guaranteed.