A Defense of the Karst Algorithm for Finding the Line of Best Fit under the L1 Norm.
Abstract
The methods proposed by Karst and Sharpe of fitting a line to a given set of points in the plane so as to minimize the sum of the absolute values of the deviations are examined by means of their linear programming formulations. For the single purpose of finding the optimal line the Karst procedure appears to be very efficient. If, as Sharpe suggests may sometimes be the case, the investigator is interested in the sensitivity of the minimum sum to changes in the slope parameter, then Sharpe's algorithm is preferred. The Karst algorithm is improved by incorporating into it the simplex stopping rule. The problem is generalized to permit arbitrary weightings of the deviations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1977
- Accession Number
- ADA042691
Entities
People
- Donald R. Schuette
Organizations
- University of Wisconsin–Madison