EEE511: Artificial Neural Computation
Systems
Last updated 11-11-2007
Final project
The difference between the final project and a practical
homework is that you are required to present the results of your project in two
ways:
- By writing a report that follows
the IEEE conference paper format. This gives you an opportunity to
practice presenting an idea in a concise and clear form and still give
sufficient detail and background for a casual but knowledgeable reader to
understand what the paper is all about. (In fact, we will be using IEEE
journal paper style files because they are available for word, too, but
they are not too far from the two-column conference paper format).
- By giving a 15-minute
presentation at the class. As above, this gives you an opportunity to
practice presenting an idea briefly to an audience that is knowledgeable
in the discipline. In this respect, it is like a presentation in a
professional conference.
Final project proposal requirements
Email the instructor a proposal (ASCII is fine) that
contains:
- A brief description of the
problem.
- A brief description of the
data that you intend to use (if this is known).
- What type of neural
network(s) do you intend to use to solve the problem and how?
- What software tools do you
intend to use?
- Your estimate of the timeline
of the project between now and end of the semester. Break the project down
into subtasks and try to estimate how much time you spend on each of
those, and make sure that it does not extend beyond the end of the
semester.
Presentation guidelines.
- You have only a 15 minute
time slot (max 12 minutes to present your work, min 3 minutes for
discussion)
- Prepare a few viewgraphs that
describe and illustrate the main points of the problem and your solution.
12 viewgraphs for 12 minutes is too fast.
- ppt, pdf, PostScript, html
are all acceptable formats for your viewgraphs
- Bring in your presentation in
a memory stick, CD, place it in an accessible place in the web, or email
to the instructor. Be early in the class on the day of your presentation
so that we can make sure your presentation is on instructor’s
computer.
Report
- Style files for the report
can be found at the IEEE
web site. Both LaTeX and MS Word style files are included.
- Maximum of six pages, four is
ideal.
- Structure of the paper:
- Abstract of four sentences
(see guidelines below).
- Motivate the problem.
- Discuss previous work, or
background of the problem.
- Present your approach, and
its details.
- Present simulation results.
- Evaluate what you have done
(how well it works and why, or why not, compare to other work in the
literature).
- Discuss further ideas and
future work (what you should/could have done given enough time)
- Use figures and graphs to
illustrate the text.
- Here is an extract from guidelines how to write a
good paper for a conference (original site here).
- Here is another guide to writing
good papers in machine learning (also as a postscript
file).
Presentation
Schedule
The last 4-5 class periods will be dedicated to the presentations. This is the schedule.