"... In PAGE 6: ... The generalized relations can be examined by users or experts interactively to filter out trivial rules and preserve interesting ones [23]. Table3 represents a generalized relation consisting of five tuples. Further generalization is needed to reduce the number of tuples.... ..."

"... In PAGE 6: ... The generalized relations can be examined by users or experts interactively to filter out trivial rules and preserve interesting ones [23]. Table3 represents a generalized relation consisting of five tuples. Further generalization is needed to reduce the number of tuples.... ..."

"... In PAGE 6: ... The generalized relations can be examined by users or experts interactively to filter out trivial rules and preserve interesting ones [23]. Table3 represents a generalized relation consisting of five tuples. Further generalization is needed to reduce the number of tuples.... ..."

"... In PAGE 7: ...Table 6. A generalized relation As shown in Table6 , different classes may share tuples. The tuples shared by different classes are called overlapping tuples.... In PAGE 8: ... sq After marking the third tuple in the class of gradu- ate and the fourth tuple in the class of undergrad., the tar- get class contains four unmarked tuples as shown in Table6 , which implies that the resulting rule will contain four disjuncts. Suppose the threshold value is 3, further generalization is performed on attribute quot;Birth_Place quot;, which results in the relation shown in Table 7.... ..."

"... In PAGE 7: ...Table 6. A generalized relation As shown in Table6 , different classes may share tuples. The tuples shared by different classes are called overlapping tuples.... In PAGE 8: ... a0 After marking the third tuple in the class of gradu- ate and the fourth tuple in the class of undergrad., the tar- get class contains four unmarked tuples as shown in Table6 , which implies that the resulting rule will contain four disjuncts. Suppose the threshold value is 3, further generalization is performed on attribute quot;Birth_Place quot;, which results in the relation shown in Table 7.... ..."

"... In PAGE 7: ...Table 6. A generalized relation As shown in Table6 , different classes may share tuples. The tuples shared by different classes are called overlapping tuples.... In PAGE 8: ... a0 After marking the third tuple in the class of gradu- ate and the fourth tuple in the class of undergrad., the tar- get class contains four unmarked tuples as shown in Table6 , which implies that the resulting rule will contain four disjuncts. Suppose the threshold value is 3, further generalization is performed on attribute quot;Birth_Place quot;, which results in the relation shown in Table 7.... ..."

"... In PAGE 19: ... The Davidon-Fletcher-Powell algorithm [9] was used for min- imizing the energy function; with a simple gradient descent the results are slightly worse for the Gaussian potential and much worse for the Lorentzian (see also Table 1). The relative generalized variance Vrel of all methods is shown in Table1 . As mentioned above, the results of the elastic algorithms (EAA and ETA) strongly depend on the minimization algorithm.... In PAGE 20: ... We also show the results of a plain least-squares estimator, implemented as a Kalman lter (KF). The last column of Table1 shows the relative computing time trel. In order to have a meaningful comparison the annealing schedules are the same in all cases, with the exception of the GSF which does not require annealing.... ..."

"... In PAGE 26: ... In the next step we have classi ed the track candidates according to their noise content. Table4 shows the range of the four classes, the number of tracks in each class, and the relative precision of the estimated track parameters, as measured by the generalized variance of the residuals with respect to the truth values. The baseline is now the class with the lowest noise content.... ..."