Accession Number : ADA062774


Title :   An Approach to the Programming of Biased Regression Algorithms.


Descriptive Note : Interim rept.,


Corporate Author : SOUTHERN METHODIST UNIV DALLAS TEX DEPT OF STATISTICS


Personal Author(s) : Gunst,Richard F


Full Text : http://www.dtic.mil/get-tr-doc/pdf?AD=ADA062774


Report Date : Nov 1978


Pagination or Media Count : 11


Abstract : Due to the near nonexistence of computer algorithms for calculating estimators and ancillary statistics that are needed for biased regression methodologies, many users of these methodologies are forced to write their own programs. Brute-force coding of such programs can result in a great waste of computer core and computing time, as well as inefficient and inaccurate computing techniques. This article proposes some guides to more efficient programming by taking advantage of mathematical similarities among several of the more popular biased regression estimators.


Descriptors :   *COMPUTER PROGRAMS , *LINEAR REGRESSION ANALYSIS , ALGORITHMS , INPUT , PROBLEM SOLVING , ESTIMATES , DIAGNOSIS(GENERAL) , BIAS


Subject Categories : Statistics and Probability
      Computer Programming and Software


Distribution Statement : APPROVED FOR PUBLIC RELEASE