Bill Langford

Voice: (805) 892-2522
FAX: (805) 892-2510

Office: Room 331

Current Research

My primary research interest is in biological and geological applications of machine learning and optimization at all scales, from molecular to global.

Some projects and interests while at NCEAS

• Evaluation functions for image segmentation
• Landscape ecology
  • Accuracy assessment in landscape ecology (with Sarah Gergel, University of British Columbia)
  • Landscape metrics for segmentation evaluation
• Cost-sensitive learning for rapid bioassessment
  • Learning to Predict Channel Stability using Biogeomorphic Features (with Stephanie Moret, Lousiana State University and Dragos Margineantu, Boeing Co.)
• Graph theory in biology and ecology
  • I coordinate the NCEAS group on graph theory in ecology.
  • Automatic learning of ontologies from textbooks
• Animal Behavior
  • Fish schooling (with Julia Parrish, University of Washington)

Past Studies

• PhD: Learning evaluation functions for 2D textured image segmentation and classification, particularly aerial images in forest, coastal, and desert environments;
• MS: Classification context in a machine learning approach to prediction of protein secondary structure

Links

• Work
• Personal