A Novel Cohesitive Implicative Classification based on MGK and application on diagnostic on informatics literacy of students of higher education in Madagascar.
Résumé
As the literature shows, a data mining is generally based on contingency tables that is on crossing properties variables; it measures statistically the quality of this rule based on nonsignificant number of counterexamples, where rule not verified.We must therefore be able to quantify the significance of these numbers, the consistency of association rule (AR) or classes of rules, the contribution of subjects or categories of subjects to some rules that represent chains of rules and, by a hierarchy, rules on rules that are also called meta-rules.