RAD
Registry Anomaly Detection

Part of the IDS research project at Columbia University.



People

Frank Apap - fsa3@columbia.edu
Andrew Honig - arh21@columbia.edu



Project Abstract

We present a host-based intrusion detection system for Microsoft
Windows. algorithm detects attacks on a host machine by looking
for anomalous accesses to the Windows Registry. The key idea is to
first train a model of normal registry behavior for a host and to
use this model to detect abnormal registry accesses at run-time.
The system trains a normal model using data that contains no attacks
and then at run-time checks each access to the registry in real
time to determine whether or not the behavior is abnormal and
corresponds to an attack. We evaluate the system by training the
system on a set of normal registry accesses and then use the system
to detect the actions of malicious software.

Project Status

We are currently polishing our paper (pdf, ps).
We plan on implementing some new features to our training and detection
algorithm. We are also setting up a demo for the RAD system.

Screenshots

Soon we will show you some pictures of the model generator and anomaly
detector in action.

Project Data

As discussed in our paper we are publishing the training data we used to build our models.
This data may be useful to someone who is also doing research on the Windows Registry.
The data is available here



Powerpoint presentation for RAID