Information Technology Reference
In-Depth Information
Table 5.1 (continued)
(a) (b) (c) Minimal cover algorithm All rules on examples algorithm
(d) (e) (f) (g) (h) (d) (e) (f) (g) (h)
21 ? cost 65 4 19 75.00 15,315 31 92 90.00
22 and gain 65 2 47 68.33 17,279 31 92 90.00
23 to cost 71 2 43 60.00 23,865 31 94 90.00
24 on cost 66 4 14 60.00 30,503 36 44 88.33
25 ; gain 62 2 36 63.33 43,096 36 44 88.33
Columns present parameters: (a) selection stage equal to the number of considered characteristic features, (b) the most recently selected variable, (c) preference
order for the most recently selected attribute, (d) number of rules in a decision algorithm without any constraints, (e) number of exact rules when they are fewer
than the total number, (f) minimal support required of rules resulting in maximal classification accuracy, (g) number of rules meeting constraints on support, (h)
maximal predictive accuracy of the classifier (%)
 
Search WWH ::




Custom Search