CSCI 1550/2540: Probabilistic Methods in Computer Science
Final Project and Review Guidelines
Spring 2024
Objectives:
• Submit a research project on any application of probabilistic methods in computer science, preferably
related to the content of the course but not mandatory.
• Give a 20 minute (15 for presentation, 5 for questions) presentation on your project in class.
• Submit short reviews for 3 out of the 9 final projects in this class.
1 Project Report
Instructions:
1. Project report is due on April 18th at 11:59 pm EST.
2. Each team must submit a single .pdf file with the NeurIPS template written in LaTeX.
3. The main paper should be 6 − 8 pages long with unlimited number of pages for appendices and
references. Any content beyond the main paper is viewed at the discretion of the reviewer.
4. Anyothersupplementarymaterials(suchasacoderepository)shouldbeincludedasalinkwithinthe
submission.
5. Optional: YoumaysubmitadraftreportbyApril 8th, 11:59 pm ESTtogetsomeearlycomments
from the instructor and the TAs.
2 In-Class Presentation
Instructions:
1. The in-class presentations will occur on April 23rd, April 25th, and April 30th.
2. Each day, 3 teams will be presenting their projects. We will release a sign-up sheet separately from
this document.
3. Each team must email their slides to the instructor and the HTA before 2:20 pm EST of their
respective presentation dates.
4. Exceptions must be requested in advance.
3 Grading of Final Project:
For Final Projects: AllstudentswhowishtosubmitafinalprojectareenrolledinCSCI2540,sothefinal
project will be scored out of 50 points.
1
CSCI 1550/2540 - Probabilistic Methods in Computer Science Page 2 of 3
1. We will run a peer-review process to score your final projects. Every student will be asked to submit
a review for 3 projects assigned to them. The instructor and the TAs will also be reviewers.
2. Each reviewer will give you a score out of 50 points. Please see Section 4 (Review) below for the
specific criterion they will use to judge your work.
3. Your total score will be the weighted average of everyone’s review. The weight of the TAs will be
twice as much as your review. The weight of the instructor’s review will be up to the discretion of the
instructor himself.
4 Review
Instructions:
1. The review is due on May 6th, 11:59 PM EST.
2. For each of the project you review, we ask you to write a short review (1 page).
Thereviewwillbeginwithasummarydescribingtheauthors’workandcontributions. Thenthereview
will give an evaluation based on the following 10 questions:
• How relevant is the project relevant to the probabilistic methods and/or computer science com-
munity?
• How original and creative are the idea and the execution?
• Are the claims and method in the report correct and rigorous to your knowledge?
• Did the authors address any limitations of their work?
• Does the project show a deep understanding of the relevant research areas/topics?
• Is it clear from the submissions which parts are the authors’ contributions?
• Is the paper well-written?
• Is the presentation well-organized?
• Does the presentation make you want to learn more about the project?
• Have the authors addressed the broader impact (including ethical and societal implications) of
their work?
Foreachofthese10questions, wealsoaskyoutoindicateascorefrom0to5onyourresponseonhow
effectively you think the authors met the criterion in each question.
Finally, the review should end with a list of typos / minor errors / other comments on the submission.
3. If you have a conflict of interests with any project groups and we happened to assign you that group,
please let us know in advance so that we can assign a different group to you.
4. If you are submitting a final project with another team member, you still need to submit the reviews
individually.
5. You may not discuss or disclose your reviews to other students in this class until the course has
ended.
6. Mandatory attendence and audience engagement for the in-class presentation are expected for
all 3 days.
CSCI 1550/2540 - Probabilistic Methods in Computer Science Page 3 of 3
5 Grading For Reviewers:
For students who will be submitting a final project, their reviews will be graded out of 10 points. For
studentswhoarenotsubmittingfinalprojects,theirreviewswillbegradedoutof30points. Thepercentage
of the grading is divided as follows:
1. (20%) Mandatory attendance and audience engagement to the in-class presentations.
2. (10%) If you are in the bottom 20% of the most accurate reviewers (measured using the least square
differences), you will receive a 0% out of 10% here. If you are not in the bottom 20% of the reviewers,
you will get all 10%.
3. (70%) The instructor and the TAs will also be reviewing your reviews. You will be evaluated on the
following criteria:
• Did the reviewer accurately describe the contributions of the authors?
• Did the reviewer adequately address all 10 questions listed in Section 4 (Review)?
• How did the reviewer provide constructive feedbacks to the submission?
6 Extra Credit:
1. Top Projects: If your project is among the top 3 projects reviewed, you will receive 3 bonus points
(in the sense of 3 out of 100 for the class).
2. Top Reviewers: If your reviews are among the top 20% of the most accurate evaluations (measured
using the least square differences), you will receive 3 bonus points.
3. Thetopprojectsandtopreviewerswillbelistedonthepubliccoursewebsiteafterthefinalgradesare
out. You can ask for your project and/or name to not be included.
