III] LADTree Classifiers:
A least absolute deviation (LAD) is used to find the error
criterion to obtain regression trees. Logical analysis of data
is one other classification method proposed in optimization
literature.In LAD a classifier is build based on learning a
logical expression. LAD is binary classifier and hence can
distinguish between positive and negative samples. The
basic assumption of LAD model is that a binary point
covered by some positive patterns, but not covered by any
negative pattern is positive, and similarly, a binary point
covered by some negative patterns, but not covered by
positive pattern is negative. For a given data set LAD
model constructs large set patterns and selects subset of
them which satisfies the above assumption such that each
pattern in the model satisfies certain requirement in terms
of prevalence and homogeneity.