Social and technological network

Projects

Final Submission deadline: November 14, 4pm. Each student must submit an Individual report. (Instructions to be uploaded on this page.)

Project submission.

The submission should include:

Report Individual report per person. pdf, 3 pages max, at least 11pt font. Optionally with an additional appendix that will not be marked.
Code The code you have implemented, along with scripts that run the code and plot the results.
Dataset The dataset. The actual data may be too large to submit, in which case put your data in a publicly readable folder and make sure your program reads from there. Or, link to a public source and include clear easy to find instructions of how to download and use the data.
(Optionally) Examples and ipython notebook. Example of running the different experiments to get the important plots.

How to submit

Create a zipped folder with all your files and name it firstname_lastname_studentID.zip. If you have large datasets it is ok to give a link to it and instructions on how to run your program.
Upload your zip file at the learn page for the course. Go to Assessment > Assignemnt submission and follow instructions. This student guidance may be helpful.
Submit early and resubmit if necessary. Don't keep it for the last minute.

Report

The report should be a 3 pages pdf document. The list of references can be outside these 3 pages. Shorter and concise is better! Make sure to describe everything in your own words. Remember to put a suiatble title to your project and mention your team. The report should include:

Abstract. A very short (few lines) abstract that states what your project is about and what you have achieved. Be precise about the main contribution/observation fromm your work.
Introduction/motivation of work. Once again, be brief. No need to start with "network analysis is important..."; get to the point! What is the big challenge in your work, what is new, what is important, and what insight/observation you have from having done the project.
Related work. Short coverage of what has been usually done in the domain before and how your work is different/similar.
Model and Problem statement. As formal/mathematical as possible. Discuss how well you think your approach solves the problem.
Your Solution/approach. What are the main indeas/insights in your approach? Why should it work?
Results. Descirbe the most important/insightful results. Show plots/figures (should be readable). What is interesting about them? How well do they fit with what you set out to do?
Appendix. Additional plots, results, discussions. You are not marked on this, but can refer to it from the main report as supporting evidence, like any other citation.

Take care with the writing. Your marks are determined by the report. So it should be clear what you set out to do, and what your interesting conributions are. The marker reading the report was not there with you on the project. So state in terms that they can understand. The most common reason for loss of marks is lack of clarity in what problem is being solved, why it is relevant and what the main ideas and results/observations are.

Project plan/proposal

1 Page Plan/Proposal deadline: October 21, any time. Ths is not marked. Feedback only. (paassed)

Submit a half page (excluding references) document concisely stating the following:

Title of your project (Think of a good one!)
Team: names and students ids
Topics and motivations. Why your project is great.
Problem statement
- Model (e.g. what is the network, what attributes are known for nodes, edges etc. what are your assumptions)
- Objective: precise formal statement of what your algorithm or heuristic or analysis hopes to achieve in this model
Background data and related works: What related papers you have found. How your work is different. What dataset(s) you have found and do they contain the relevant attributes? (or can you derive them somehow?)
Preliminary ideas: how do you hope to approach it? (e.g. what type of techniues you might use.) Consider looking up more techniques from last year's course. See slides used in Oct 11 lecture.

On Piazza, post a note in the project folder, with the proposal as a pdf attachment. You are free to post it publicly, in which case other classmates can contribute to the discussion.

Put an appropriate title for the note. In the body of the post, mention your team, and any specific questions or discussion points you have. We will try to answer them.

About the project

The coursework project is intended to give you a chance to try out your ideas.

Every year, students in the STN course have fun exploring new ideas and playing with different datasets. Some typical types of projects are:

Given a dataset from some domain, devise a simple prediction or recommendation method
Find interesting properties of specific networks
Design an efficient algorithm for a network analysis problem
Use network analysis to improve certain ML tasks.

Depending on a student's interest, a good project can later be converted into a research paper. In the coursework itself, a clear explanation of the idea and a few experimental plots suffices.

One-two weeks after the start of project, you can submit a short description and get feedback on your project concept and progress.

To give you a better idea, the following are a few sample projects from recent editions of the course:

You can see other examples at: the similar course taught at Stanford.