Hailfinder: A Bayesian system for forecasting severe weather

https://doi.org/10.1016/0169-2070(95)00664-8Get rights and content

Abstract

Hailfinder is a Bayesian system that combines meteorological data and model with expert judgment, based on both experience and physical understanding, to forecast severe weather in Northeastern Colorado. The system is based on a model, known as a belief network (BN), that has recently emerged as the basis of some powerful intelligent systems. Hailfinder is the first such system to apply these Bayesian models in the realm of meteorology, a field that has served as the basis of many past investigations of probabilistic forecasting. The design of Hailfinder provides a variety of insights to designers of other BN-based systems, regardless of their fields of application.

References (20)

  • B. Abramson et al.

    Probabilistic forecasts from probabilistic models

  • B. Abramson et al.

    Using belief networks to forecast oil prices

    International Journal of Forecasting

    (1991)
  • B. Abramson et al.

    Towards an art and science of knowledge engineering

  • K. Drake et al.

    Demos Tutorial-An Introduction to Demos

    (1991)
  • W. Edwards et al.

    Bayesian statistical inference for psychological research

    Psychological Review

    (1963)
  • D.E. Heckerman

    Probabilistic Similarity Networks

    (1991)
  • D.E. Heckerman et al.

    Toward normative expert systems: Part I. The pathfinder project

    Methods of Information in Medicine

    (1992)
  • M. Henrion et al.

    Decision analysis and expert systems

    AI Magazine

    (1991)
  • S. Holtzman

    Intelligent Decision Systems

    (1989)
  • F.V. Jensen

    An Introduction to Bayesian Networks

    (1996)
There are more references available in the full text version of this article.

Cited by (150)

  • A decomposition-based algorithm for learning the structure of multivariate regression chain graphs

    2021, International Journal of Approximate Reasoning
    Citation Excerpt :

    ALARM [5] with 37 vertices, 46 edges and 509 parameters, it was designed by medical experts to provide an alarm message system for intensive care unit patients based on the output a number of vital signs monitoring devices. HAILFINDER [1] with 56 vertices, 66 edges, and 2656 parameters, it was designed to forecast severe summer hail in northeastern Colorado. The natural continuation of the work presented here would be to develop a learning algorithm with weaker assumptions than the one presented.

  • Knowledge Verification From Data

    2024, IEEE Transactions on Neural Networks and Learning Systems
  • General power laws of the causalities in the causal Bayesian networks

    2024, International Journal of General Systems
View all citing articles on Scopus

This work was supported in party by the National Science Foundation under grants SES-9106440 and IRI-9424378, in part by the Forecast Systems Laboratory of the National Oceanic and Atmospheric Administration, and in part by the Wood Kalb Foundation.

View full text