Commit a3798ab6 authored by Robert Ricci's avatar Robert Ricci

The talk stuff is going to be Lecture 21 this year

parent b91ccc0a
......@@ -10,9 +10,9 @@
\newfontfamily\titlefont[Numbers=OldStyle,Ligatures=TeX]{Equity Caps A}
\allsectionsfont{\titlefont}
\title{CS6963 Lecture \#19}
\title{CS6963 Lecture \#17}
\author{Robert Ricci}
\date{March 17, 2014}
\date{April 8, 2014}
\begin{document}
......@@ -20,25 +20,55 @@
\begin{outline}
\1 Figuring out what the factors are that caused us to get different results. 2d grid:
\2 Results: DM, Reno w/o, Reno w/, Cubic (SACK doesn;t matter)
\2 Methods: Runs (Few/lots), method of launching clients (threads, client-side, server-side), controlling time (number of runs, size of fetch)
\2 Goal here is not to declare someone right and wrong, goal is to understand under what conditions we get different results
\1 Review:
\2 One run
\2 Three runs, five runs
\2 Turn on display of variance
\2 Turn display of mean CI
\2 Pick the number of runs
\2 Approximate test
\1 First half of a good report
\2 Intro clearly states goals---even though they are in the plan, it helps focus the report
\2 Followed procedure to pick the number of runs --- but did he actually re-calc confidence for all of them
\2 TODO: Figures 1 and 2
\2 Regional network conditions - one table, easy to compare,
\2 Discussion as we go
\1 Post-mortem for Lab 1
\2 Was our plan sufficient?
\2 What could/should we have done differently?
\1 Grading Lab 1
\2 Everyone will try to reproduce someone else's result!
\2 Assignments to be passed out after class
\2 Due next Tuesday
\1 Giving talks - specifically, conference type talks
\1 Hold the audience's interest
\2 Tell a story
\2 Assume they are looking for an excuse to get back to their email
\2 Put in the effort: multiply the audience size by the length of your
talk to think of how much attention you are consuming
\1 Slides
\2 Your slides are not your notes
\3 They are visual aids to punctuate what you're saying
\2 If you and the slides say the same thing, one is not necessary
\2 You are competing with your slides for audience attention
\3 Don't put up complicated stuff all at once---progressive reveal
\3 Use bullet lists as little as possible
\2 Slides are not paper, you don't have to do black on white
\2 Use color to convey information, but remember
\3 Conference projectors have the worst color accuracy in the universe
\3 Colorblindness, esp. red/green colorblindness (8\% of
European-descended men)
\2 If you're not comfortable with humor, just put the joke on the slide
\2 Animations should only reduce confusion or convey information
\1 Preparation
\2 Do practice talks---and consider the audience
\2 Listen to feedback, don't argue with it
\2 Do two practices after the final change you make to your slides
\2 It's important to be comfortable (physically as well as mentally)
\2 Practice the key words that you might have trouble pronouncing
\1 Presentation
\2 Never look at your slides
\2 Never look at your slides
\2 Don't us a laser pointer---use animated objects on the slides
\2 Talk to people, not the room
\2 Use a clicker
\1 For next time
\2 Two papers to read, evals to do
\2 Work on reproducing!
\end{outline}
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment