diff options
| author | jeanpouget-abadie <jean.pougetabadie@gmail.com> | 2015-01-26 11:48:53 -0500 |
|---|---|---|
| committer | jeanpouget-abadie <jean.pougetabadie@gmail.com> | 2015-01-26 11:48:53 -0500 |
| commit | 7265b4b5ff05ec64b88ec8698724dfd5b235f29f (patch) | |
| tree | c324b0760dae3595d84f7858594c60182afdea60 /paper/sections/assumptions.tex | |
| parent | 294537d14612e2ab05e8c90362b0efde3f76675a (diff) | |
| download | cascades-7265b4b5ff05ec64b88ec8698724dfd5b235f29f.tar.gz | |
adding files
Diffstat (limited to 'paper/sections/assumptions.tex')
| -rw-r--r-- | paper/sections/assumptions.tex | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/paper/sections/assumptions.tex b/paper/sections/assumptions.tex new file mode 100644 index 0000000..eccd095 --- /dev/null +++ b/paper/sections/assumptions.tex @@ -0,0 +1,6 @@ +\subsection{The Restricted Eigenvalue Condition} + +Proving the restricted eigenvalue assumption for correlated measurements is non-trivial. Under reasonable assumptions on the graph parameters, we can show a very crude ${\cal O}(N)$-lower bound for $\gamma_n$ by exploiting only the first set of measurements, where only the source nodes are active. Note that even though we waste a lot of information, we obtain similar asymptotic behavior than previous work. + +\subsection{The Irrepresentability Condition} + |
