Abstract
It has long been popular to utilize the least Squares estimation procedure for fitting the multiple linear regression model to observed data. In this paper, two useful alternatives to least Squares (L2 norm) estimation in exploratory data analysis are examined: least absolute value estimation (L1 norm) and Chebychev (L∞ norm) estimation. Formulating the L1 norm and L∞ norm problems as linear programming problems offers several advantages, including efficient Solution methods using special-purpose Computer codes. An example is provided in which the three procedures are used to fit a line, both with and without an outlier present in the data.
Get full access to this article
View all access options for this article.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.