(Should I be more formal about these things, or is this about right?)

////////////////////////////////////////////////////////////////////////////

[To Michael Jordan]

Subject: Looking for an external Ph.D. thesis committee member

Hi Michael --

I'm a Ph.D. student of Andrew Moore's at Carnegie Mellon University
ready to propose a thesis primarily on using graphical models for
dataset compression and condensed representations.  (We met briefly
when you visited CMU a while ago at a group chat about reinforcement
learning.)  Anyhow, Andrew and I think you would make an excellent
external committee member for the thesis, partially since there's a
very good chance I'll be examining and using variational methods
throughout much of it.  If you think you might be interested, you can
download a copy of my current proposal draft at 
www.cs.cmu.edu/~scottd/proposal.ps; I'd certainly be very happy to
hear any of your thoughts on it (even if you decide not to act as the
external committee member).

Best wishes,

-- Scott

////////////////////////////////////////////////////////////////////////////

[To John Lafferty]

Subject: Looking for a Ph.D. thesis committee member

Hi John --

I'm just about ready to propose a thesis on using probabilistic models to
do dataset compression and to speed up data mining.  Anyhow, I think
you would make an excellent thesis committee member, since you're
obviously quite well-versed in information theory and statistical
learning algorithms. :)  If you think you might be interested, you
can download a copy of my current proposal draft at 
www.cs.cmu.edu/~scottd/proposal.ps; I'd certainly be very happy to
hear any of your thoughts on it (even if you decide not to act
as a committee member).

Best wishes,

-- Scott

///////////////////////////////////////////////////////////////////////////

[To Tom Mitchell]

Subject: Looking for a Ph.D. thesis committee member

Hi Tom --

I'm just about ready to propose a thesis on using machine learning to
automatically learn models from data in order to perform dataset
compression and to speed up future data mining, and Andrew and I both
think you'd make a great committee member.  If you think you might be
interested, you can download a copy of my current proposal draft at
www.cs.cmu.edu/~scottd/proposal.ps; I'd certainly be very happy to
hear any of your thoughts on it (even if you decide not to act as a
committee member)!

Best wishes,

-- Scott

////////////////////////////////////////////////////////////////////////////

[To Christos Falutsos]

Subject: Looking for a Ph.D. thesis committee member

Hi Christos --

I'm just about ready to propose a thesis primarily on using Bayesian
networks to perform dataset compression and as condensed representations.
Andrew and I both think you'd make an excellent comittee member,
since you're the departmental database guru and the whole
Bayesian-networks-as-condensed-representations theme is largely due to
my earlier discussions with you. :)  If you think you might be
interested, you can download a copy of my current proposal draft at
www.cs.cmu.edu/~scottd/proposal.ps; I'd certainly be very happy to
hear any of your thoughts on it (even if you decide not to act as a
committee member)!

Best wishes,

-- Scott