Algorithms for reasoning in credal trees

01 October 2011 → 30 September 2015
Regional and community funding: Special Research Fund, Research Foundation - Flanders (FWO)
Research disciplines
  • Natural sciences
    • Applied mathematics in specific fields
    • Statistics and numerical methods
    • Artificial intelligence
    • Computer architecture and networks
    • Distributed computing
    • Information sciences
    • Information systems
    • Programming languages
    • Scientific computing
    • Theoretical computer science
    • Visual computing
    • Other information and computing sciences
  • Social sciences
    • Cognitive science and intelligent systems
algorithms imprecise probabilities credal networks
Project description

Develop theory and efficient algorithms for inferences in credal trees, with emphasis on imprecise hidden Markov models (iHMM). These represent a system’s uncertain evolution through states, where we can only observe the states imperfectly, through uncertain outputs. I plan to adress: learning an iHMM model from sequences of observations, dealing with missing data and extending results to general credal trees.