Timeline for How Does Random Noise Typically Look?
Current License: CC BY-SA 2.5
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 24, 2009 at 15:24 | comment | added | Ben Weiss | Hi Gil, I knew I was missing something, but I guess I'm still not exactly clear what it is. Are you asking the following: There is the set of vectors in $\mathbb{F}_2^k$ of weight (number of $1$'s) no more than $t.$ Somehow you want a ``random" (for some definition of random) way of choosing one of those vectors and garbling your message by adding it? Is this more what you want? That it should be random, but correlated by not having more than $t$ errors? | |
| Nov 24, 2009 at 15:11 | comment | added | Gil Kalai | Dear Ben, Thanks! The crux of the question is if when we properly define a notion of "random noise" does it behave (when we restrict the number of garbled bits) like the noise you described, namely like a situation where the bits are garbled independently. Greg Kuperberg proposed to define random noise via random stochastic maps, see his answer. | |
| Nov 24, 2009 at 14:57 | history | answered | Ben Weiss | CC BY-SA 2.5 |