Credit points: 3
Grading policy:
10% - presence && participation
10% - paper reviews (to be submitted until 8am Sunday to the lecturer and to the course staff by e-mail, in text or pdf ( NO MS WORD please)
80% - presentation ( please discuss with Mark - possible via e-mail - no later than Thursday the week before you present )
Review guidelines:
Your review should address the following (1 paragraph each)
1. The scope + brief summary of contribution
2. Pros and cons of the approach, hidden problems and potential limitations, what you liked or disliked about the paper (.e.g. the paper is trying to solve the problem X on GPU Y, but it won't scale, the experiments do not prove the claim, and the solution cannot be applied to some other cases, and it basically reinvents the wheel introduced in the work by Author Z from 20 years ago in a more general context)
3. Open questions, possible research directions
Presentation guidelines:
1. Provide a comprehensive overview of the subject, and touch related work with the emphasis on the contribution of the paper(s) being presented
2. Provide an in-depth description of the most important issues the paper solves. Don't try to cover all the small details, give a big picture, and then delve into the most important parts.
3. Be assertive, devil's advocate, show the weaknesses and the strengths
4. Suggest possible future directions, open issues.
5. Plan for 20 slides/per hour to allow discussion
Grading policy:
10% - presence && participation
10% - paper reviews (to be submitted until 8am Sunday to the lecturer and to the course staff by e-mail, in text or pdf ( NO MS WORD please)
80% - presentation ( please discuss with Mark - possible via e-mail - no later than Thursday the week before you present )
Review guidelines:
Your review should address the following (1 paragraph each)
1. The scope + brief summary of contribution
2. Pros and cons of the approach, hidden problems and potential limitations, what you liked or disliked about the paper (.e.g. the paper is trying to solve the problem X on GPU Y, but it won't scale, the experiments do not prove the claim, and the solution cannot be applied to some other cases, and it basically reinvents the wheel introduced in the work by Author Z from 20 years ago in a more general context)
3. Open questions, possible research directions
Presentation guidelines:
1. Provide a comprehensive overview of the subject, and touch related work with the emphasis on the contribution of the paper(s) being presented
2. Provide an in-depth description of the most important issues the paper solves. Don't try to cover all the small details, give a big picture, and then delve into the most important parts.
3. Be assertive, devil's advocate, show the weaknesses and the strengths
4. Suggest possible future directions, open issues.
5. Plan for 20 slides/per hour to allow discussion