CSE 416a Final Project (FL19)

This is an inactive course webpage.

 

Scope

The purpose of the final project is to do network analysis that builds on the things we learned in class and goes beyond that in some aspect, such as exploring novel

  • real-word applications,
  • algorithms,
  • efficient implementations,
  • analysis techniques,
  • network models, or
  • creating your own network (analysis problem and dataset).
TOPICS
  • FEATURED Topics
    • Fairness: Homophily and the Glass Ceiling Effect in Social Networks: paper
    • Crazy idea: Build, Run, and Organize your own Social Network: HowTo
    • Motif Algorithms: Building blocks of biological networks: a review on major network motif discovery algorithms: paper
    • CNA in Biology: Evolution of resilience in protein interactomes across the tree of life: paper
    • CNA in Biology, Node Classification: Graphlet Kernels for Prediction of Functional Residues in Protein Structures: main paper, potentially useful paper
    • Graph-based Machine Learning, Graph Classification: Efficient graphlet kernels for large graph comparison: paper
  • For other ideas, check the resources page sections on Datasets and Code, as well as, Papers & Ideas for Final Project.
TEAMS
  • team sizes are 2-3 students (not more – not less)
  • each team member should take responsibility for a distinct component of the project
  • we will reserve the right to fuse or split teams based on the submitted proposals
Phase 1: PROPOSALS
  • due THU 10/31 at 11:59pm (plus automatic extension)
  • use the proposal template
    • use the Abstract for the proposal, then you can turn in into an Abstract fort he final report
    • this is a LaTeX template – if you haven’t used LaTeX before consider using Overleaf
      • easy to use – check these tutorials
      • online – no need to install LaTeX on your computer!
      • collaborative – share the report with your team members and work not he same document (it has version control!)
  • proposal specifications
    • what do you want to do in your project?
    • why?
    • how will you do this? Short description!
    • what are the resources?
      • data
      • implementations/software packages
      • literature/paper(s)
      • include references (links, citations, etc.)
    • write at max 400 words
    • add three keywords (need to be more specific than “Network Analysis”)
  • submit pdf via Gradescope (fp_proposal)
Phase 2: PROPOSAL FEEDBACk
  • available FRI 11/08
  • teams may be fused or split based on the proposal content
  • we reserve the right to reject proposals (in that case the team will be dissolved or encouraged to submit a new proposal)
Phase 3: Work on Project
  • we expect each team to put in the same amount of effort as for 2.5 homework assignments
  • start working on the project early
  • make sure you understand the data you want to work with
  • make sure you understand the algorithms you want to work with (even though if they are available as NetworkX/SNAP functions!)
  • VISUALIZATION: Use network visualization (e.g., Gephi – an open source graph visualization and analysis tool) for both data understanding and to present/explain results. You will have to try different layouts and experiment with node and edge coloring, sizes, etc.! Add careful descriptions on what your visualizations show.
FINAL PRESENTATIONS
  • TUE 12/03 and THU 12/05 in-class
  • each team will get a 10-15min time slot (depending on the number of teams)
  • IMPORTANT:
    • 15min means
      • 12min for presentation
        • you cannot speak longer than 12min
        • this includes setup and transition time
      • 3min for questions
    • 20min (for combines teams) means
      • 15min for both presentations
        • you cannot speak longer than 15min
        • this includes setup and transition time
      • 5min for questions to both teams
  • project presentations schedule and slide submission instructions: here
  • every team member will need to be actively taking part in the team’s presentation
  • peer feedback will be required
    • each student will be assigned to 4 presentations to give feedback
    • feedback form will be linked from HERE
SHORT REPORT
  • due: FRI 12/13 at 8pm (no extension)
  • use your proposal template
  • 4-5 pages (can be exceeded by 1-2 pages for figures only)
  • include your (updated) proposal as an abstract
  • focus on (prettynetwork visualizations and result tables, plots, figures
  • code will have to be accessible to instructor via public or private repo (provide url in your project report)
  • submit report via Gradescope (fp_report)
GRADING
  • 15% team proposal (instructor/TA evaluated)
  • 25% team presentation (based on instructor/TA and peer feedback)
  • 10% for giving good/constructive feedback on presentations (individual)
  • 50% team report and implementation

Your project report should include a motivation, description of approach (methodology), a detailed description of the experimental results, as well as a discussion of your results. Your report should be self-contained and understandable for an informed outsider. It should not require the reader to look at any of your code, to run it, nor to read the reference paper(s).
The grading of the report will heavily account for cleanliness (structure), readability (logical flow, writing style), and presentation (figures, tables, plots, etc.)!

Example ProjectS

Here are two example projects from cse416a in Fall 2018: