ON ASYMPTOTIC SIMULTANEOUS CONFIDENCE REGIONS FOR REGRESSION SURFACES IN A CURVILINEAR POISSON REGRESSION MODEL WITH AN APPLICATION IN ROAD SAFETY RESEARCH

The problem of finding simulataneous confidence regions for regression surfaces is considered in a curvilinear Poisson regression model. A general procedure is given for constructing simultaneous confidence intervals for all linear combinations of the mean vector of a normally distributed random vector with known covariance matrix. The problem of finding all square roots of a given covariance matrix is studied. Asymptotic simultaneous confidence regions of different types for regression surfaces are constructed. The regression model and the theory developed are applied to the study of effects of speed limits on the number of accidents. The number of vehicle kilometres driven on the total road net outside built-up areas per time unit is considered to be the independent variable./TRRL/

  • Corporate Authors:

    Stockholm University, Sweden

    P.O. Box 6701
    S-11385 Stockholm,   Sweden 
  • Authors:
    • Haglund, T
  • Publication Date: 1973

Media Info

  • Features: Figures; References; Tables;
  • Pagination: 92 p.

Subject/Index Terms

Filing Info

  • Accession Number: 00126339
  • Record Type: Publication
  • Source Agency: Transport and Road Research Laboratory (TRRL)
  • Report/Paper Numbers: NR 73 R&D Rept.
  • Files: ITRD, TRIS
  • Created Date: Apr 21 1976 12:00AM