Invitation to Participate, Call for Papers and Contributions


In conjunction with The Eleventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD-2005,
Chicago, IL - August, 2005

Workshop Description

    Over the last decade there have been many advances in automated construction of predictive models from partially or fully labeled data collected from data streams, single or distributed databases, sensor data, and unsupervised discovery of regularities in data. In the meantime, the emphasis on prediction has meant little effort has been devoted to other important aspects of data analysis. In particular, there has been relatively little work on principled methods for discovering anomalies or outliers in data. Specifically, these are irregularities that cannot be explained by existing domain models or knowledge.

    Research efforts in the area of anomaly detection is spread across a number of communities (security, health, biology, environmental sciences, manufacturing, business, economics, and finance), and much of the existing work focuses on detecting outliers solely for the purpose of removing them from the analysis to prevent them from unduly affecting the data mining process instead of treating them as interesting phenomena in their own right.

    This workshop will be a forum for presenting and discussing principled approaches for anomaly discovery in large scale data sets, outlier detection in static images, video, and multi-media data, fast detection of anomalies in time-sensitive tasks, anomalies in high-risk applications (such as security), distinguishing between relevant and less relevant outliers, trade-offs in tuning algorithms for anomaly detection, integration of anomaly detection systems and human experts, and for assessing the capabilities of anomaly detection for data mining and learning methods.

    Topics of interest include, but are not limited to:

    - Anomaly detection in spatio-temporal data
    - Anomaly detection based on multiple data sources
    - Integration of data mining components and expert knowledge for anomaly detection
    - Analysis of the capabilities of learning algorithms for anomaly detection
    - Automated anomaly detection in safety-critical applications
    - Algorithms and tools for online outlier detection
    - Limiting and reducing false alarm rates; analysis of error tradeoffs
    - Validation and testing of anomaly detection systems, metrics for their performance
    - Failures in applying data mining and learning on anomaly detection
    - Novel application domains
    - Testbeds for the evaluation and testing of anomaly detection methods.
    The workshop also aims to explore the common tasks that need to be addressed in practical applications that require anomaly detection tools and algorithms, such as data collection, sampling, and pre-processing. Organizing the workshop in conjunction with KDD 2005 will provide a unique opportunity for bringing together researchers working on fundamental questions regarding algorithms and basic technology, and practitioners who work or have interests in addressing different anomaly detection tasks by employing data mining technologies.

    The workshop is intended to be a follow-up of the Symposium on Machine Learning for Anomaly Detection that has taken place in May 2004, at Stanford University ( - but with a more heavy emphasis on data mining and knowledge discovery techniques, and open to all researchers who express interest in the topic of anomaly detection.

Workshop Format
    The workshop will have two sessions. Each session will start with an invited talk and will continue as a mix of position paper presentations and discussions.

Participation and Submissions
    To participate in the workshop, please send an e-mail message to Dragos Margineantu ( giving your name, affiliation, address, e-mail address, and a brief description of your reasons for wanting to attend.
    In addition, if you wish to present a position paper on one or more of the topics listed above, please see the instructions on the submissions page.

    If you have an issue or contribution that is not covered by the topics above, please contact Dragos Margineantu by e-mail to discuss your idea prior to submitting a position paper. The organizers and the program committee will review the submissions with the goal of assembling a stimulating and exciting workshop. Attendence will be limited to 40-50 people, with preference given to people who are presenting position papers.

Important dates
  • Submission deadline: June 10, 2005.
  • Notification of acceptance: July 1, 2004.
  • Workshop to be held on August 21, 2005.

  • Dragos Margineantu, Boeing, Mathematics & Computing Technology
  • Stephen Bay, PricewaterhouseCoopers
  • Philip Chan, Florida Institute of Technology
  • Terran Lane, University of New Mexico