diff options
| author | Stratis Ioannidis <stratis@stratis-Latitude-E6320.(none)> | 2012-11-01 23:42:57 -0700 |
|---|---|---|
| committer | Stratis Ioannidis <stratis@stratis-Latitude-E6320.(none)> | 2012-11-01 23:42:57 -0700 |
| commit | b6aa0b124de02bda659db226ef66f1886f98252e (patch) | |
| tree | 9c2caa7b4bc139819a04b14adfe42fdabf51749d /general.tex | |
| parent | b041a0d77807e18d02e266e1a0aca057db803973 (diff) | |
| download | recommendation-b6aa0b124de02bda659db226ef66f1886f98252e.tar.gz | |
stratis comments
Diffstat (limited to 'general.tex')
| -rw-r--r-- | general.tex | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/general.tex b/general.tex index ea66aed..530b8d3 100644 --- a/general.tex +++ b/general.tex @@ -22,6 +22,7 @@ Our objective \eqref{,,,} clearly follows from \eqref{bayesianobjective} by sett Moreover, our results can be extended to the general Bayesian case, by replacing $I_d$ with the positive semidefinite matrix $R$: \thibaut{Discussion about the value function below} +\stratis{The text below is unpolished/not written for external consumption. Rather than trying to motivate $R>I$, it is probably better to figure out a lower bound.} When there is an $R$ in the value function, it seems to make more sense to study the modified value function: |