ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Here is the template you are required to use for submission: https://www.overleaf.com/read/btddtdztdrgs
2
Please let us know if you have trouble accessing the overleaf template.
3
4
To begin the proposal writing part, here are some tips:
5
6
Keep your proposal concise:
You are trying to pitch an idea and you should convey your thoughts as comprehensively as possible but while being concise!
Try to limit to 2 pages unless you have done some significant work, or have an extensive literature review (which I highly support that you start working on!).
The more papers you read the better chances you have to come up with novel ideas that can be published.
7
8
Include the following sections / footnotes:
9
Introduction
10
Literature Review
11
Baseline model ideas from research papers that you have read
12
(Optional: Methods IF you already have started working on models.)
13
Dataset description
14
Division of Work (e.g., "Author A designed the study, Author B performed the experiments...")
15
- include division of work as a footnote or endnote to the paper
16
17
Maintain a Github Repository:
If you have also started the modeling process (not expected at this stage, but some students like to start early), begin a Github repository and include the link in your proposal.
It will be required for the final report.
18
19
A Project can be...
20
A novel idea/an interesting take on an existing chain of work and you want to implement & experiment with it yourself to see how it performs.
This is the standard project direction and is structured very similarly to a standard research project. The field can be anything from language to CV to RL, etc.
21
Reproduction of a conference paper to verify the core results of the paper. This may involve implementing the paper yourself if the code is not available.
Further, if the code is available, this is not simply an exercise in re-running an open sourced code base -- you should develop experiments that
provide a means to question critical sections/takeaways of a paper and prove/disprove their findings. The difficulty in this direction is no less than that of developing a project yourself.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
-- TAs and Prof
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100