[TYPES/announce] Postdoctoral position on graph data mining and anomaly detection

James Cheney james.cheney at gmail.com
Tue May 30 17:19:37 EDT 2017


Dear types/announce,

I am advertising a postdoctoral position on graph data mining and anomaly
detection, with applications to security.  I realize this may seem like an
off-the-wall topic to advertise here, but thought I would try anyway in
case there is anyone out there with a background in probabilistic
programming or PL techniques for data science who might like to try out
such techniques on a new application area.  Further details below.

--James

I am pleased to announce that we are now accepting applications for a
postdoctoral research position in graph data mining and anomaly detection.
The position is for 18 months, starting on or around September 1, 2017 and
by January 1, 2018 at the latest.  Funding is provided by a DARPA grant on
the project: "ADAPT: A diagnostics approach to advanced persistent threat
detection", part of the $60M Transparent Computing program.

The ADAPT project is led by Galois, Inc. in collaboration with Oregon State
University and the University of Edinburgh.  The overall aim of the project
is to manage and analyze large / high-volume streams of provenance data
recording low-level operating system activity in order to detect small,
"anomalous" subgraphs that may comprise system attacks.  This is a
challenging problem because available attack data is sparse and unlabeled,
and future attackers can be expected to mimic normal system behavior and
avoid previous attack patterns.

This postdoctoral position will contribute to applying existing anomaly
detection techniques to this new domain, identifying the most useful
techniques for this data or developing new, better-suited approaches, and
integrating them into an analysis and visualization pipeline to aid
security investigation.

Applications must be received by 5pm GMT, June 27, 2016.  Applications must
include a current CV and short (1-3 page) statement of research interests
and their relevance to the project.  To apply, visit the University job
posting for this position:

https://www.vacancies.ed.ac.uk/pls/corehrrecruit/erq_
jobspec_version_4.jobspec?p_id=039859

then click "apply" and follow the instructions.  Please note that
applicants must use the University's application system above, which
involves some account registration and form-filling, and it is recommended
that applicants complete this process well before the deadline, since the
system automatically stops accepting applications after the deadline.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://LISTS.SEAS.UPENN.EDU/pipermail/types-announce/attachments/20170531/99fc1dce/attachment.html>


More information about the Types-announce mailing list