AAAI 2000 Workshop Schedule


Learning from Imbalanced Data Sets

Workshop Schedule

8:30-9:30: Session 1: Theoretical and Practical Issues

  • 8:30-9:15: Foster Provost (Invited Talk), "Machine Learning from Imbalanced Data Sets 101"

  • 9:15-9:30: David Stork, "Open Mind Animals: Insuring the Quality of Data Openly Contributed over the World Wide Web"

9:30-9:45: Discussion

9:45-10:15: Session 2: Empirical Studies

  • 9:45-10:00: Nathalie Japkowicz, "Learning from Imbalanced Data Sets: A Comparison of Various Strategies"

  • 10:00-10:15: Pamela Surko, "Correlates of State Failure"

10:15-10:30: Discussion

10:30-11:00: MORNING BREAK

11:00-12:15: Session 3: Recognition-Based Approaches

  • 11:00-11:30: Stephen Muggleton (Invited Talk), "Measuring Performance when Positives are Rare"

  • 11:30-11:45: David Tax, "Feature Scaling in Support Vector Data Descriptions"

  • 11:45-12:00: Vadim Bulitko, "Using Autoencoding Networks for Tramp Metal Detection"

12:00-12:15: Discussion

12:15-12:30: Morning Wrap-up

12:30-2:00: LUNCH BREAK

2:00-3:15: Session 4: Biasing the Inductive Process

  • 2:00-2:30: Rich Caruana (Invited Talk), "Learning from Imbalanced Data: Rank Metrics and Extra Tasks"

  • 2:30-2:45: Ed Pednault, "Imbalanced Data Sets in Insurance Risk Modelling"

  • 2:45-3:00: Gary Weiss, "Learning to Predict Extremely Rare Events"

  • 3:00-3:15: Jerzy Grzymala-Busse, "An Approach to Imbalanced Data Sets Based on Changing Rule Strength"

3:15-3:30: Discussion

3:30-4:00: AFTERNOON BREAK

4:00-4:30: Session 5: Imbalanced Data Sets and Cost Sensitivity

  • 4:00-4:15: Robert Holte, "Exploiting the Cost (In)sensitivity of Decision Tree Splitting Criteria"

  • 4:15-4:30: Dragos Margineantu, "When Does Imbalanced Data Require More than Cost Sensitive Learning?"

4:30-4:45: Discussion

4:45-5:00: Afternoon Wrap-up

5:00-5:30: DISCUSSION, PUBLICATION PLANS and CONCLUDING REMARKS, Rob Holte, Stan Matwin