Project Proposal
Project Proposal Instructions
Follow the same guidelines (presented below) for both the presentation and the report.
We will be very strict w.r.t the timing of your presentation slides. Its going to be 15 mins (with 10 mins of presentation + 5 minutes of QA). Once again, since there are 8 teams presenting in a class, we will not tolerate going overtime. Please practice your presentation.
Report due 3/12 (Tuesday, 11:59 pm ET). MAX 4 pages excluding references, and one submission per team (submit on gradescope). Report template: Link.
Please include the following in your project proposal:
Title for your project (this can also change later on). Something descriptive but also memorable.
Introduction section which must contain the following subsections:
Task / Research Question Description: What is the task you are trying to solve or what is the research question you are trying to answer?
Motivation & Limitations of existing work: Have others tried to solve the same task or answer a similar research question? What are you trying to do differently and why? What were the limitations or shortcomings of prior work?
Proposed Approach: Briefly describe some of your initial thoughts about your proposed approach and your preliminary ideas
Likely challenges and mitigations: What is hard about this task / research question? What are your contingency plans if things turn out to be harder than expected or experiments do not go as planned?
Related Work Section
Include 2-3 sentence descriptions of no less than 4 relevant papers. Also mention how your work differs from these.
Experiments section
Datasets - Please list which datasets you plan to use , whether or not you have access them, and whether or not they are publicly available with the same preprocessing and train / dev / tests as the previous work you will be comparing to (if applicable).
Evaluation Metrics - Please explain how you will measure progress towards achieving your task. Justify why this is a suitable approach, and discuss potential limitations of the evaluation.
Baselines - Please list the baseline methods (if applicable) that you will compare to, importantly state whether or not you will need to run their method on new datasets or if you have already published results. Also list if they have publicly available source code that can be used to run experiments.
Software - Please list some of the software libraries / components / existing code that you will be building your project off of.
Timeline - Please provide a week-by -week timeline for the next few weeks leading up to the midpoint presentation. What experiments will you do each week? What would you like to have accomplished by the midterm? How will you implement an end-to-end If you're in a team how will you divide up the work?
Challenges - Lay down the challenges that you can foresee which might prevent you from attaining your goal.
Contingency Plan - Set of backup experiments / model that you would try out in case if the original plan does not work out.