Network Modeling for the Internet
Semi-Experiments - Formal Testing
Network Modeling for the Internet
Semi-Experiments - Formal Testing
| Leader | Mike Devetsikiotis (North
Carolina State University), mdevets@ncsu.edu
http://www4.ncsu.edu/~mdevets |
| Meeting | Tuesday 1:00 - 2:00 pm, room 104 -- jointly with "Semi-Experiments: Look & See" group |
| Members | Ian Dinwoodie (Duke University),
ihd@stat.duke.edu
Arka Ghosh (University of North Carolina, Chapel Hill), apghosh@email.unc.edu Steve Marron (SAMSI/University of North Carolina, Chapel Hill), marron@email.unc.edu Jonathan Mattingly (Duke University), jonm@math.duke.edu George Michailidis (University of Michigan), gmichail@umich.edu David Rolls (SAMSI), rollsd@uncw.edu Zhengyuan Zhu (University of North Carolina, Chapel Hill), zhuz@email.unc.edu |
| Outline Objectives | The task of systematically
characterizing the multitude of "factors" affecting traffic behavior and
network performance, is a difficult but central one. Assessing the "effects"
of such factors, ranging from qualitative statements, to full quantification,
to (even better) sensitivity analysis, is particularly important. The approach
of "semi-experiments" seems to present a refreshingly novel approach to
this line of work.
The objective of this work group is to engage in the careful and systematic study of such "semi-experiments" and attempt to (a) formalize them in a statistical sense (i.e., relate them to design of experiments, factorial analysis, statistical confidence, hypothesis testing, etc.); and (b) combine them or extend them utilizing the group's own expertise in statistics, output analysis techniques, and network system design methodologies. |
| References | Semi-Experiments and the Modelling of Internet Traffic (by Darryl Veitch) |
Meetings
10/22/03: Discussion on Metrics
11/04/03: Queuing behavior
12/16/2003: George's
five minute presentation
© 2002, Statistical and Applied Mathematical Sciences Institute. All rights reserved.