A workshop organized in association with ICDM’2017 November 18, 2017 New Orleans, USA
Modern industrial and infrastructure systems have a cyber-physical architecture, incorporating the underlying physical processes regulated by a control cyber system. In other words, they represent interdependent networks described by disparate mathematical models creating scientific challenges that go well beyond the modeling and analysis of the individual system layers. Moreover, this complex structure is at the origin of vulnerabilities of the system to internal and external failures, as well as to the cyber attacks. Therefore, the questions of development of adequate models and methods for an efficient detection and localization of intrusions and faults, as well as proportional responses to potential attacks is of prime significance in many real-world industrial systems, such as automotive systems, smart manufacturing, power grid, HVAC and building systems, etc.
Given that the detailed information about the underlying system topology and interactions might not be available, the emerging data mining techniques are becoming fundamental to tackle the aforementioned challenges. These techniques requires ideas and methodology from a wide variety of fields, including but not limited to statistical modeling, graph theory, anomaly detection, optimization, time series analytics, etc. Focusing on the methodological and practical aspects of data mining for industrial systems, this workshop provides an opportunity to discuss the latest theoretical advances and real-world applications in the field of cyber-physical systems. Papers are solicited to address a wide range of topics in these areas, including but not limited to:
Data Mining Methods for Industrial Systems
- Anomaly detection
- Correlation discovery
- Intrusion detection, localization and identification
- Model selection and online learning of data streams
- Custering and dimensionality reduction
- Network-based analysis
Applications and Testbeds
- Modeling of cyber-physical systems: avionics, automotive, advanced manufacturing, buildings, HVAC systems, smart grids, transportation, health care
- Cyber security for infrastructures
- Fault detection in industrial systems
- Control and SCADA systems
- Unique industrial data sets
Note that the workshop presentations will run in a joint session "Data Mining in Cyber" with the workshop "Machine Learning in Cyber" focusing on theoretical machine learning aspects for cyber security.
- August 7, 2017: Due date for full workshop papers
- September 7, 2017: Notification of workshop papers acceptance to authors
- September 15, 2017: Camera-ready deadline for accepted papers
- November 18, 2017: Workshop date
To appear soon
Accepted papers will be included in the IEEE ICDM 2017 Workshops Proceedings volume published by IEEE Computer Society Press, and will also be included in the IEEE Xplore Digital Library. The workshop proceedings will be in a CD separated from the CD of the main conference. The CD is produced by IEEE Conference Publishing Services (CPS). Every workshop paper must have at least one paid registration in order to be published.
- Scott Backhaus (Los Alamos National Laboratory)
- Aric Hagberg (Los Alamos National Laboratory)
- Nathan Lemons (Los Alamos National Laboratory)
- Andrey Lokhov (Los Alamos National Laboratory)
To appear soon