AbstractWe study discrete L_1, curve-fitting of n points in k dimensional space. Execution times for the algorithms of Barrodaie-Roberts (BR), Bartels-Conn-Sinclair (BCS), and Bloomfieid-Steiger (BS) - three of the best LAD curve-fitting procedures - are compared over a variety of deterministic and random curve-fitting problems. Analysis of the results allows us to make surprisingly precise statements about the computational complexity of these algorithms. In particular, BR is in a different complexity ciass than BCS and BS as the number of points,n, increases. All algorithms are linear in'the dimension, k, andBS is less complex than BCS.
SubjectsLeast absolute deviation, Curve-fitting, Linear programming
RightsThis Item is protected by copyright and/or related rights.You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use.For other uses you need to obtain permission from the rights-holder(s).